REVIEWARTICLE

Nasal drug delivery devices: characteristics and performancein a clinical perspective—a review

Per Gisle Djupesland

Published online: 18 October 2012# The Author(s) 2012. This article is published with open access at Springerlink.com

Abstract Nasal delivery is the logical choice for topicaltreatment of local diseases in the nose and paranasal sinusessuch as allergic and non-allergic rhinitis and sinusitis. Thenose is also considered an attractive route for needle-freevaccination and for systemic drug delivery, especially whenrapid absorption and effect are desired. In addition, nasaldelivery may help address issues related to poor bioavail-ability, slow absorption, drug degradation, and adverseevents in the gastrointestinal tract and avoids the first-passmetabolism in the liver. However, when considering nasaldelivery devices and mechanisms, it is important to keep inmind that the prime purpose of the nasal airway is to protectthe delicate lungs from hazardous exposures, not to serve asa delivery route for drugs and vaccines. The narrow nasalvalve and the complex convoluted nasal geometry with itsdynamic cyclic physiological changes provide efficient fil-tration and conditioning of the inspired air, enhance olfac-tion, and optimize gas exchange and fluid retention duringexhalation. However, the potential hurdles these functionalfeatures impose on efficient nasal drug delivery are oftenignored. With this background, the advantages and limita-tions of existing and emerging nasal delivery devices anddispersion technologies are reviewed with focus on theirclinical performance. The role and limitations of the in vitrotesting in the FDA guidance for nasal spray pumps andpressurized aerosols (pressurized metered-dose inhalers)with local action are discussed. Moreover, the predictivevalue and clinical utility of nasal cast studies and computersimulations of nasal airflow and deposition with computer fluiddynamics software are briefly discussed. New and emergingdelivery technologies and devices with emphasis on Bi-Directional™ delivery, a novel concept for nasal delivery that

can be adapted to a variety of dispersion technologies, aredescribed in more depth.

Keywords Drug delivery . Nasal . Device . Paranasalsinuses . Topical . Systemic . Vaccine . Nasal valve .

Particle deposition . Clearance

Introduction

Intuitively, the nose offers easy access to a large mucosalsurface well suited for drug- and vaccine delivery. However,factors related to the nasal anatomy, physiology and aero-dynamics that can severely limit this potential, have histor-ically been challenging to address. The most recent FDAguidance for nasal devices provides detailed guidelines forin vitro testing of the physical properties such as in vitroreproducibility and accuracy of plume characteristics anddose uniformity of mechanical liquid spray pumps andpressurized metered-dose inhalers (pMDIs) for nasal use[1]. The guidance primarily addresses in vitro testing ofnasal sprays and pressurized aerosols for local action. Thereference to in vivo performance is limited to the recom-mendation of minimizing the fraction of respirable particlesbelow 9 μm in order to avoid lung inhalation of drugsintended for nasal delivery. Thus, although important asmeasures of the quality and reliability of the spray pumpand pMDI mechanics, these in vitro tests do not necessarilypredict the in vivo particle deposition, absorption, and clin-ical response [2]. Furthermore, the guidance offers no orlimited guidance on nasal products for systemic absorptionand for alternative dispensing methods like drops, liquidjets, nebulized aerosol, vapors, and powder formulations.Finally, it does not address aspects and challenges related tothe nasal anatomy and physiology that are highly relevantfor the device performance in the clinical setting like body

P. G. Djupesland (*)OptiNose,Oslo, Norwaye-mail: pgd@optinose.no

Drug Deliv. and Transl. Res. (2013) 3:42–62DOI 10.1007/s13346-012-0108-9

position, need for coordination, and impact of airflow andbreathing patterns at delivery.

The mechanical properties of different modes of aerosolgeneration are already well described in depth in a previouspublication [3]. The anatomy and physiology of the nasalairway has also recently been summarized in an excellentrecent review [4]. The aim of this paper is to take a stepfurther by reviewing the characteristics of existing andemerging nasal delivery devices and concepts of aerosolgeneration from the perspective of achieving the clinicalpromise of nasal drug and vaccine delivery. Focus is put ondescribing how the nasal anatomy and physiology presentsubstantial obstacles to efficient delivery, but also on how itmay be possible to overcome these hurdles by innovativeapproaches that permit realization of the therapeutic potentialof nasal drug delivery. Specific attention is given to theparticular challenge of targeted delivery of drugs to the uppernarrow parts of the complex nasal passages housing themiddle meatus where the sinuses openings are located, aswell as the regions innervated by the olfactory nerve andbranches of the trigeminal nerve considered essential forefficient “nose-to-brain” (N2B) transport.

Nasal anatomy and physiology influencing drug delivery

Regulation of nasal airflow

Nasal breathing is vital for most animals and also for humanneonates in the first weeks of life. The nose is the normaland preferred airway during sleep, rest, and mild exercise upto an air volume of 20–30 l/min [5]. It is only when exercisebecomes more intense and air exchange demands increasethat oral breathing supplements nasal breathing. The switchfrom nasal to oronasal breathing in young adults appearswhen ventilation is increased to about 35 l/min, about fourtimes resting ventilation [6]. More than 12,000 l of air passthrough the nose every day [5]. The functionality of the noseis achieved by its complex structure and aerodynamics.Amazingly, the relatively short air-path in the nose accountsfor as much as 50–75 % of the total airway resistance duringinhalation [7, 8].

The nasal valve and aerodynamics

The narrow anterior triangular dynamic segment of the nasalanatomy called the nasal valve is the primary flow-limitingsegment, and extends anterior and posterior to the head ofthe inferior turbinate approximately 2–3 cm from the nostrilopening [9]. This narrow triangular-shaped slit acts as adynamic valve to modify the rate and direction of theairflow during respiration [10, 11]. Anatomical studies de-scribe the static valve dimensions as 0.3–0.4 cm2 on each

side, whereas acoustic rhinometry studies report the func-tional cross-sectional area perpendicular to the acousticpathway to be between 0.5 and 0.6 cm2 on each side, inhealthy adults, with no, or minimal gender differences[11–14]. The flow rate during tidal breathing creates airvelocities at gale force (18 m/s) and can approach the speedof a hurricane (32 m/s) at sniffing [11, 15]. At nasal flowrates found during rest (up to 15 l/min), the flow regimen ispredominantly laminar throughout the nasal passages. Whenthe rate increases to 25 l/min, local turbulence occurs down-stream of the nasal valve [10, 11, 15]. The dimensions canexpand to increase airflow by dilator muscular action knownas flaring, or artificially by mechanical expansion by inter-nal or external dilators [16, 17]. During inhalation, Bernoulliforces narrow the valve progressively with increasing inspi-ratory flow rate and may even cause complete collapse withvigorous sniffing in some subjects [5]. During exhalation, thevalve acts as a “brake” to maintain a positive expiratoryairway pressure that helps keep the pharyngeal and lowerairways open and increase the duration of the expiratoryphase. This “braking” allows more time for gas exchange inthe alveoli and for retention of fluid and heat from the warmsaturated expiratory air [4, 17, 18]. In fact, external dilation ofnarrow noses in obstructive sleep apnea patients had benefi-cial effects, whereas dilation of normal noses to “supernor-mal” dimensions had deleterious effects on sleep parameters[17]. However, in the context of nasal drug delivery, the smalldimensions of the nasal valve, and its triangular shape thatnarrows further during nasal inhalation, represent importantobstacles for efficient nasal drug delivery.

The nasal mucosa—filtration and clearance

The region anterior to the valve called the vestibule is lined bynon-ciliated squamous epithelium that in the valve regiongradually transitions into ciliated epithelium typical of theciliated respiratory epithelium posterior to the valve region[4, 19]. Beyond the nasal valve, the nasal turbinates divide thenasal cavity into slit-like passages with much larger cross-sectional area and surface area (Figs. 1, 2 and 3). Here, thepredominantly laminar airflow is slowed down to speeds of2–3 m/s and disrupted with eddies promoting deposition ofparticles carried with the air at and just beyond the valveregion [11]. The ciliated respiratory mucosa posterior to thenasal valve is covered by a protective mucous blanketdesigned to trap particles and microorganisms [4, 19].Thebeating action of cilia moves the mucous blanket towardsthe nasopharynx at an average speed of 6 mm/min (3–25 mm/min) [20, 21]. The large surface area and close contactenables effective filtering and conditioning of the inspired airand retention of water during exhalation (Figs. 1, 2 and 3). Oralbreathing increases the net loss of water by as much as 42 %compared to nasal breathing [22]. The nasal passages were

Drug Deliv. and Transl. Res. (2013) 3:42–62 43

optimized during evolution to protect the lower airways fromthe constant exposure to airborne pathogens and particles.

Specifically, particles larger than 3–10 μm are efficientlyfiltered out and trapped by the mucus blanket [19]. The nosealso acts as an efficient “gas mask” removing more that 99 %of water-soluble, tissue-damaging gas like sulfur dioxide [23].Infective agents are presented to the abundant nasal immunesystem both in the mucous blanket, in the mucosa, and in theadjacent organized lymphatic structures making the nose at-tractive for vaccine delivery with potential for a longstandingcombination of systemic and mucosal immune responses [24].The highly vascularized respiratory mucosa found beyond thevalve allows exchange of heat and moisture with the inspiredair within fractions of a second, to transform cold winter airinto conditions more reminiscent of a tropical summer [19].

The nasal cycle

The physiological alternating congestion and decongestionobserved in at least 80 % of healthy humans is called the nasalcycle [5, 25]. The nasal cycle was first described in therhinological literature by a German physician in 1895, butwas recognized in Yoga literature centuries before [5]. Healthyindividuals are normally unaware of the spontaneous andirregular reciprocal 1–4-h cycling of the nasal caliber of thetwo individual passages, as the total nasal resistance remainsfairly constant [26]. The autonomic cyclic change in airflowresistance is mainly dependent on the blood content of thesubmucosal capacitance vessels that constitute the erectilecomponent at critical sites, notably the nasal valve region.Furthermore, the erectile tissues of the septal and lateral wallsand the turbinates respond to a variety of stimuli includingphysical and sexual activity and emotional states that canmodify and override the basic cyclic rhythm [4]. The cycleis present during sleep, but overridden by pressures applied tothe lateral body surface during recumbency to decongest theuppermost/contralateral nasal passage. It has been suggestedthat this phenomenon causes a person to turn from one side tothe other while sleeping [5, 27]. The cycle is suppressed inintubated subjects, but restored by resumption of normal nasalbreathing [28]. The cycle may also cause accumulation ofnitric oxide (NO) in the congested passage and adjacentsinuses and contribute to defense against microbes throughdirect antimicrobial action and enhanced mucociliary clear-ance [29]. Measurements have shown that the concentrationof NO in the inspired air is relatively constant due to theincrease in NO concentration within the more congested cav-ity, which nearly exactly counterbalances the decrease in nasalairflow [30]. In some patients, as a result of structural devia-tions and inflammatory mucosal swelling, the nasal cycle maybecome clinically evident and cause symptomatic obstruction[19]. Due to the cycle, one of the nostrils is considerably morecongested than the other most of the time, and the vastmajority of the airflow passes through one nostril while theother remains quite narrow especially at the valve region [5].

Fig. 1 The complex anatomy of the nasal airways and paranasal sinuses

Fig. 2 Illustration of the breath-powered Bi-Directional™ technology.See text for detailed description

44 Drug Deliv. and Transl. Res. (2013) 3:42–62

Consequently, the nasal cycle contributes significantly to thedynamics and resistance in the nasal valve region and must betaken into consideration when the efficiency of nasal drugdelivery devices is considered.

Nasal and sinus vasculature and lymphatic system

For nasally delivered substances, the site of deposition mayinfluence the extent and route of absorption along with thetarget organ distribution. Branches of the ophthalmic andmaxillary arteries supply the mucous membranes coveringthe sinuses, turbinates, meatuses, and septum, whereas thesuperior labial branch of the facial artery supplies the part ofthe septum in the region of the vestibule. The turbinateslocated at the lateral nasal wall are highly vascularized witha very high blood flow and act as a radiator to the airway.They contain erectile tissues and arteriovenous anastomosesthat allow shunting and pooling related to temperature andwater control and are largely responsible for the mucosalcongestion and decongestion in health and disease [19, 31].

Substances absorbed from the anterior regions are morelikely to drain via the jugular veins, whereas drugs absorbedfrom the mucosa beyond the nasal valve are more likely todrain via veins that travel to the sinus cavernous, where thevenous blood comes in direct contact with the walls of thecarotid artery. A substance absorbed from the nasal cavity tothese veins/venous sinuses will be outside the blood–brainbarrier (BBB), but for substances such as midazolam, whicheasily bypass the BBB, this route of local “counter-currenttransfer” from venous blood may provide a faster and moredirect route to the brain. Studies in rats support that a preferen-tial, first-pass distribution to the brain through this mechanismafter nasal administration may exist for some, but not all smallmolecules [32, 33]. The authors suggested that this counter-current transport takes place in the area of the cavernous sinus–carotid artery complex, which has a similar structure in rat andman, but the significance of this mechanism for nasally deliv-ered drugs has not been demonstrated in man [32, 33].

The lymphatic drainage follows a similar pattern as thevenous drainage where lymphatic vessels from the vestibuledrain to the external nose to submandibular lymph nodes,whereas the more posterior parts of the nose and paranasalsinuses drain towards the nasopharynx and internal deeplymph nodes [4]. In the context of nasal drug delivery,perivascular spaces along the olfactory and trigeminalnerves acting as lymphatic pathways between the CNS andthe nose have been implicated in the transport of moleculesfrom the nasal cavity to the CNS [34].

Innervation of the nasal mucosa

The nose is also a delicate and advanced sensory organdesigned to provide us with the greatest pleasures, but also

to warn and protect us against dangers. An intact sense ofsmell plays an important role in both social and sexualinteractions and is essential for quality of life. The senseof smell also greatly contributes to taste sensations [35].Taste qualities are greatly refined by odor sensations, andwithout the rich spectrum of scents, dining and wining andlife in general would become dull [36]. The olfactory nervesenter the nose through the cribriform plate and extenddownwards on the lateral and medial side of the olfactorycleft. Recent biopsy studies in healthy adults suggest that theolfactory nerves extend at least 1–2 cm further anterior anddownwards than the 8–10 mm described in most textbooks(see Figs. 1 and 2) [37, 38]. The density decreases, butolfactory filaments and islets with olfactory epithelium arefound in both the anterior and posterior parts at the middleturbinate. In addition, sensory fibers of both the ophthalmicand maxillary branches of the trigeminal nerve contribute toolfaction by mediating a “common chemical sense” [39].Branches of the ophthalmic branch of the trigeminal nerveprovide sensory innervation to the anterior part of the noseincluding the vestibule, whereas maxillary branches inner-vate the posterior part of the nose as well as the regions witholfactory epithelium.

The olfactory and trigeminal nerves mutually interact in acomplex manner. The trigeminal system can modulate theolfactory receptor activity through local peptide release orvia reflex mechanisms designed to minimize the exposure toand effects of potentially noxious substances [39]. This canoccur by alteration of the nasal patency and airflow andthrough changes in the properties of the mucous blanketcovering the epithelium. Trigeminal input may amplifyodorous sensation through perception of nasal airflow andat the chemosensory level. Interestingly, an area of increasedtrigeminal chemosensitivity is found in the anterior part ofthe nose, mediating touch, pressure, temperature, and pain[39]. Pain receptors in the nose are not covered by squamousepithelium, which gives chemical stimuli almost direct ac-cess to the free nerve endings. In fact, loss of trigeminalsensitivity and function, and not just olfactory nerve func-tion, may severely reduce the sense of smell [40]. Thisshould not be forgotten when addressing potential causesof reduced or altered olfaction.

The sensitivity of the nasal mucosa as a limiting factor

In addition to the limited access, obstacles imposed by its smalldimensions and dynamics, the high sensitivity of the mucosa inthe vestibule and in the valve area is very relevant to nasal drugdelivery. Direct contact of the tip of the spray nozzle duringactuation, in combination with localized concentrated anteriordrug deposition on the septum,may create mechanical irritationand injury to the mucosa resulting in nosebleeds and crusting,and potentially erosions or perforation [41]. Furthermore, the

Drug Deliv. and Transl. Res. (2013) 3:42–62 45

high-speed impaction and low temperature of some pressurizeddevices may cause unpleasant sensations reducing patient ac-ceptance and compliance.

The role of the high sensitivity of the nasal mucosa as anatural nasal defense is too often neglected when the poten-tial of nasal drug delivery is discussed, in particular whenresults from animal studies, cast studies, and computer fluiddynamics (CFD) are evaluated. Exposure to chemicals, gas-es, particles, temperature and pressure changes, as well asdirect tactile stimuli, may cause irritation, secretion, tearing,itching, sneezing, and severe pain [39]. Sensory, motor, andparasympathetic nerves are involved in a number of nasalreflexes with relevance to nasal drug delivery [4]. Suchsensory inputs and related reflexes are suppressed by theanesthesia and/or sedation often applied to laboratory ani-mals, potentially limiting the clinical predictive value ofsuch studies. Further, the lack of sensory feedback andabsence of interaction between the device and human sub-jects/patients are important limitations of in vitro testing ofairflow and deposition patterns in nasal casts and in CFDsimulation of deposition. Consequently, deposition studiesin nasal casts and CFD simulation of airflow and depositionare of value, but their predictive value for the clinical settingare all too often overestimated.

Targeted nasal delivery

For most purposes, a broad distribution of the drug on themucosal surfaces appears desirable for drugs intended forlocal action or systemic absorption and for vaccines [3].However, in chronic sinusitis and nasal polyposis, targeteddelivery to the middle and superior meatuses where thesinus openings are, and where the polyps originate, appearsdesirable [42, 43]. Another exception may be drugs intendedfor “nose-to-brain” delivery, where more targeted deliveryto the upper parts of the nose housing the olfactory nerveshas been believed to be essential. However, recent animaldata suggest that some degree of transport can also occuralong the branches of the first and second divisions of thetrigeminal nerve innervating most of the mucosa at andbeyond the nasal valve [44]. This suggests that, in contrastto the prevailing opinion, a combination of targeted deliveryto the olfactory region and a broad distribution to the mu-cosa innervated by the trigeminal nerve may be optimal forN2B delivery. Targeted delivery will be discussed in moredetail below.

Nasal drug delivery devices

The details and principles of the mechanics of particlegeneration for the different types of nasal aerosols havebeen described in detail by Vidgren and Kublik [3] in their

comprehensive review from 1998 and will only be brieflydescribed here, with focus instead on technological featuresdirectly impacting particle deposition and on new andemerging technologies and devices. Liquid formulationscurrently completely dominate the nasal drug market, butnasal powder formulations and devices do exist, and moreare in development. Table 1 provides an overview of themain types of liquid and powder delivery devices, their keycharacteristics, and examples of some key marketed nasalproducts and emerging devices and drug–device combina-tion products in clinical development (Table 1).

Devices for liquid formulations

The liquid nasal formulations are mainly aqueous solutions,but suspensions and emulsions can also be delivered. Liquidformulations are considered convenient particularly for top-ical indications where humidification counteracts the dry-ness and crusting often accompanying chronic nasaldiseases [3]. In traditional spray pump systems, preserva-tives are typically required to maintain microbiological sta-bility in liquid formulations. Studies in tissue cultures andanimals have suggested that preservatives, like benzalko-nium chloride in particular, could cause irritation and re-duced ciliary movement. However, more recent humanstudies based on long-term and extensive clinical use haveconcluded that the use of benzalkonium chloride is safe andwell tolerated for chronic use [45]. For some liquid formu-lations, in particular peptides and proteins, limited stabilityof dissolved drug may represent a challenge [46].

Drops delivered with pipette

Drops and vapor delivery are probably the oldest forms ofnasal delivery. Dripping breast milk has been used to treatnasal congestion in infants, vapors of menthol or similarsubstances were used to wake people that have fainted, andboth drops and vapors still exist on the market (e.g.,www.vicks.com). Drops were originally administered bysucking liquid into a glass dropper, inserting the dropperinto the nostril with an extended neck before squeezing therubber top to emit the drops. For multi-use purposes, dropshave to a large extent been replaced by metered-dose spraypumps, but inexpensive single-dose pipettes produced by“blow-fill-seal” technique are still common for OTC prod-ucts like decongestants and saline. An advantage is thatpreservatives are not required. In addition, due to inadequateclinical efficacy of spray pumps in patients with nasal pol-yps, a nasal drop formulation of fluticasone in single-dosepipettes was introduced in the EU for the treatment of nasalpolyps. The rationale for this form of delivery is to improvedrug deposition to the middle meatus where the polypsemerge [47, 48]. However, although drops work well for

46 Drug Deliv. and Transl. Res. (2013) 3:42–62

Table 1 Overview of the main types of liquid and powder delivery devices, their key characteristics, and examples of some key marketed nasalproducts and emerging devices and drug–device combination products in clinical development

LIQUID DEVICESProduct stage

Example substance(s)

Example use/indication(s) Dosing Mechanism Actuation

ReferencesRelevant website(s)

Vapor

Vapor inhaler Marketed Methol Rhinitis, Common cold Multidose VaporisationNasal inhalation www.vicks.com

Drops

Rhinyle catheter (Ferring)Marketed product(s) Desmopressin Diabetes incipidus Single dose Mechanical

Breath powered 50,51,114 www.ferring.com

Multi-dose droppers (multiple)Marketedproduct(s) Decongestants Rhinitis, Common cold Multi-dose Mechanical Hand actuated

Unit-dose pipettes (multiple)Marketed product(s) Topical steroids Nasal polyps Single dose Mechanical Hand actuated 47,48 www.gsk.com

Mechanical spray pumpsSqueeze bottle (multiple)

Marketed product(s) Decongestants Rhinitis, Common cold Multi-dose Mechanical Hand actuated 3

www.novarits.com ; www.afrin.com

Multi-dose Metered-dose spray pumps (multiple) 3,13,14www,aptar.com ; www.rexam.com

Marketed product(s) Topicals steroids

Allergic & Perinneal rhinitis Multi-dose Mechanical Hand actuated 45,52

www.gsk.com www.merck.com ; www.az.com

Marketed product(s) Desmopressin

Primary nocturnal enuresis Multi-dose Mechanical Hand actuated 50,51,114 www.ferring.com

Marketed product(s) Calcitonin Osteoporosis Multi-dose Mechanical Hand actuated 3 www.novartis.comMarketedproduct(s) Ketrorolac Pain Multi-dose Mechanical Hand actuated 3 www.luitpold.com Marketed product(s) Oxytocin

Induction of lactation & labor Multi-dose Mechanical Hand actuated 3 www.defiante.com

Single/duo-dose spray pumps (multiple) 3Marketed product(s) Triptans

Migraine & Cluster headache Single dose Mechanical Hand actuated 58,59

www.gsk.com ; www.az.com

Marketed device Device Vaccine, CRS Single dose Mechanical Hand actuated 53,54 www.lmana.comMarketed device Vaccine Influenza vaccine Single dose Mechanical Hand actuated 55,56 www.bd.comMarketed device Vaccines Undisclosed

Single/duo-dose Mechanical Hand actuated 57 www.crucell.com

Bi-dir Multi-dose spray pump (OptiNose)

Product in Phase 3

Fluticasone propionate CRS with Nasal polyps Multi-dose Mechanical

Breath powered

13,50,87,88,91,95 www.optinose.com

Gas driven spray systems/atomizersSlow spray HFA pMDI's (Teva/3M)

Marketed product(s) Topicals steroids Allergic rhinitis Multi-dose Gas propellant Hand actuated 60-62

www.teva.com ; www.3m.com

Nitrogen gass driven (Impel) Preclinical Not known Not known Multi-dose Gas driven Gas driven 77www.impelneuropharma.com

Electrically powered Nebulizers/Atomizers Pulsation membrane nebulizer (Pari)

Marketed device Topical steroids

Sinusitis and nasal polyps Multi-dose Electical Electical 68 www.pari.com

Vibrating mech nebulizer (Aerogen)

Marketed device Topical drugs

Sinusitis and nasal polyps Multi-dose Electical Electical 72 www.aerogen.com

Hand-held mechanical nebuliser (Kurve)

Marketed device

Insulin (Phase 2 trials) Alzheimer's, Sinusitis Multi-dose Electical Electical

42, 71,73-76 www.kurvetech.com

POWDER DEVICESProduct stage Substances Indication(s) Dosing Mechanism Actuation

ReferencesExample website(s)

Mechanical powder sprayersPowder spray device (capsule based) (SNBL) Phase 2 Zlomitriptan Migraine Singel dose Mechanical Hand actuated 81,82 www.snbl.com

Powder sprays (Aptar/Vallois) Device Not known Not known Single dose Mechanical Hand actuated 84www.bd.com; www.aptar.com

Powder spray device (BD)Marketed device Not known Not known Single dose Mechanical Hand actuated 83 www.bespak.com

Powder spray device (Bespak)Marketed device Not known Not known Single dose Mechanical Hand actuated 85 www.bespak.com

Breath actuated inhalers

Multi-dosepowderinhaler(AZ)Rhinocort Turbohaler Budesonide

Allergic rhinitis, Nasal polyps Multi-dose Mechanical

Nasal inhalation 79,80 www.az.com

Single/duo dose capsule inhaler (Nippon-Shinyaku) Twin-lizer

Dexamethasone cipecilate Allergic rhinits

Single/duo dose Mechanical

Nasal inhalation

www.nippon-shinyaku.co.jp

Nasal inhlaler (Aptar/Pfeiffer) DeviceAmpomorhine (discont.) Parkinson's Single/duo Mechanical

Nasal inhalation

www.aptar.com ; www.stada.de

Insufflators

Insufflator - (Trimel) Preclinical Undisclosed Allergic rhinitis, Single dose MechanicalExhalation driven 86

www.trimelpharmaceuticals.com

Breath powered Bi-directional delivery (OptiNose) Phase 3 trials

Sumatriptan powder Migraine Single dose Mechanical

Breath powered 14,94,98,90 www.optinose.com

Drug Deliv. and Transl. Res. (2013) 3:42–62 47

some, their popularity is limited by the need for head-downbody positions and/or extreme neck extension required forthe desired gravity-driven deposition of drops [43, 49].Compliance is often poor as patients with rhinosinusitisoften experience increased headache and discomfort inhead-down positions.

Delivery of liquid with rhinyle catheter and squirt tube

A simple way for a physician or trained assistant to depositdrug in the nose is to insert the tip of a fine catheter ormicropipette to the desired area under visual control andsquirt the liquid into the desired location. This is often usedin animal studies where the animals are anesthetized orsedated, but can also be done in humans even without localanesthetics if care is taken to minimize contact with thesensitive mucosal membranes [50]. This method is, howev-er, not suitable for self-administration. Harris et al. [51]described a variant of catheter delivery where 0.2 ml of aliquid desmopressin formulation is filled into a thin plastictube with a dropper. One end of the tube is positioned in thenostril, and the drug is administered into the nose as drops oras a “liquid jet” by blowing through the other end of the thintube by the mouth [51]. Despite a rather cumbersome pro-cedure with considerable risk of variability in the dosing,desmopressin is still marketed in some countries with thisrhinyle catheter alongside a nasal spray and a tablet fortreatment of primary nocturnal enuresis, Von Willebranddisease, and diabetes insipidus.

Squeeze bottles

Squeeze bottles are mainly used to deliver some over-the-counter (OTC) products like topical decongestants. Bysqueezing a partly air-filled plastic bottle, the drug is atom-ized when delivered from a jet outlet. The dose and particlesize vary with the force applied, and when the pressure isreleased, nasal secretion and microorganisms may be suckedinto the bottle. Squeeze bottles are not recommended forchildren [3].

Metered-dose spray pumps

Metered spray pumps have, since they were introducedsome four decades ago, dominated the nasal drug deliverymarket (Table 1). The pumps typically deliver 100 μl (25–200 μl) per spray, and they offer high reproducibility of theemitted dose and plume geometry in in vitro tests. Theparticle size and plume geometry can vary within certainlimits and depend on the properties of the pump, the formu-lation, the orifice of the actuator, and the force applied [3].Traditional spray pumps replace the emitted liquid with air,and preservatives are therefore required to prevent

contamination. However, driven by the studies suggestingpossible negative effects of preservatives, pump manufac-turers have developed different spray systems that avoid theneed for preservatives. These systems use a collapsible bag, amovable piston, or a compressed gas to compensate for theemitted liquid volume [3] (www.aptar.com and www.rexam.-com). The solutions with a collapsible bag and a movablepiston compensating for the emitted liquid volume offer theadditional advantage that they can be emitted upside down,without the risk of sucking air into the dip tube and compro-mising the subsequent spray. This may be useful for someproducts where the patients are bedridden and where a head-down application is recommended. Another method used foravoiding preservatives is that the air that replaces the emittedliquid is filtered through an aseptic air filter. In addition, somesystems have a ball valve at the tip to prevent contamination ofthe liquid inside the applicator tip (www.aptar.com). Thesepreservative-free pump systems become more complex andexpensive, and since human studies suggest that preservativesare safe and well tolerated, the need for preservative-freesystems seems lower than previously anticipated [45]. Morerecently, pumps have been designed with side-actuation andintroduced for delivery of fluticasone furoate for the indicationof seasonal and perennial allergic rhinitis [52]. The pump wasdesigned with a shorter tip to avoid contact with the sensitivemucosal surfaces. New designs to reduce the need for primingand re-priming, and pumps incorporating pressure point fea-tures to improve the dose reproducibility and dose countersand lock-out mechanisms for enhanced dose control andsafety are available (www.rexam.com and www.aptar.com).Importantly, the in vivo deposition and clinical performanceof metered-dose spray pumps can be enhanced for someapplications by adapting the pumps to a novel breath-powered “Bi-Directional™” delivery technology describedin more detail below [13].

Single- and duo-dose spray devices

Metered-dose spray pumps require priming and some degreeof overfill to maintain dose conformity for the labeled numberof doses. They are well suited for drugs to be administereddaily over a prolonged duration, but due to the priming pro-cedure and limited control of dosing, they are less suited fordrugs with a narrow therapeutic window. For expensive drugsand vaccines intended for single administration or sporadicuse and where tight control of the dose and formulation is ofparticular importance, single-dose or duo-dose spray devicesare preferred (www.aptar.com).

A simple variant of a single-dose spray device (MAD) isoffered by LMA (LMA, Salt Lake City, UT, USA; www.lma-na.com). A nosepiece with a spray tip is fitted to a standardsyringe. The liquid drug to be delivered is first drawn into thesyringe and then the spray tip is fitted onto the syringe. This

48 Drug Deliv. and Transl. Res. (2013) 3:42–62

device has been used in academic studies to deliver, for exam-ple, a topical steroid in patients with chronic rhinosinusitis andin a vaccine study [53, 54]. A pre-filled device based on thesame principle for one or two doses (Accuspray™, BectonDickinson Technologies, Research Triangle Park, NC, USA;www.bdpharma.com) is used to deliver the influenza vaccineFluMist (www.flumist.com), approved for both adults and chil-dren in the US market [55, 56]. A similar device for two doseswas marketed by a Swiss company for delivery of anotherinfluenza vaccine a decade ago. This vaccine was withdrawndue to occurrence of adverse events (Bell’s palsy) potentiallyrelated to the cholera toxin adjuvant used [57]. The devicetechnology is now owned by a Dutch vaccine company (Cru-cell N.V. Leiden, the Netherlands; www.crucell.com), but to ourknowledge is not currently used in any marketed products.

The single- and duo-dose devices mentioned above con-sist of a vial, a piston, and a swirl chamber. The spray isformed when the liquid is forced out through the swirl cham-ber. These devices are held between the second and the thirdfingers with the thumb on the actuator. A pressure pointmechanism incorporated in some devices secures reproduc-ibility of the actuation force and emitted plume characteristics[58]. Currently, marketed nasal migraine drugs like Imitrex(www.gsk.com) and Zomig (www.az.com; Pfeiffer/Aptarsingle-dose device) and the marketed influenza vaccine Flu-Mist (www.flumist.com; Becton Dickinson single-dose spraydevice) are delivered with this type of device [59] (Table 1).With sterile filling, the use of preservatives is not required, butoverfill is required resulting in a waste fraction similar to themetered-dose, multi-dose sprays. To emit 100 μl, a volume of125μl is filled in the device (Pfeiffer/Aptar single-dose device)used for the intranasal migraine medications Imitrex(sumatriptan) and Zomig (zolmitriptan) and about half of thatfor a duo-dose design [58].

Nasal pressurized metered-dose inhalers (pMDIs)

Most drugs intended for local nasal action are delivered byspray pumps, but some have also been delivered as nasalaerosols produced by pMDIs. Following the ban on ozone-depleting chlorofluorocarbon (CFC) propellants, the numberof pMDI products for both pulmonary and nasal deliverydiminished rapidly, and they were removed from the USmarket in 2003 [60]. The use of the old CFC pMDIs fornasal products was limited due to complaints of nasal irri-tation and dryness. The particles from a pMDI are releasedat a high speed and the expansion of a compressed gas,which causes an uncomfortable “cold Freon effect” [61].The particles emitted from the traditional pMDIs had aparticle velocity much higher than a spray pump (5,200 vs.1,500 cm/s at a distance 1–2 cm from the actuator tip) [3].The issues related to the high particle speed and “cold Freoneffect” have been reduced with the recently introduced

hydrofluoroalkane (HFA)-based pMDI for nasal use offer-ing lower particle speeds [60]. Recently, the first nasalpMDI using HFA as propellant to deliver the first generationtopical steroid beclomethasone dipropionate (BDP) was ap-proved for allergic rhinitis in the USA [62]. Like spraypumps, nasal pMDIs produce a localized deposition on theanterior non-ciliated epithelium of the nasal vestibule and inthe anterior parts of the narrow nasal valve, but due to quickevaporation of the spray delivered with a pMDI, noticeable“drip-out” may be less of an issue [63].

Mismatch between geometry of anterior nose and the sprayplume

The pressure created by the force actuating a spray pumpdrives the liquid through the swirl chamber at the tip of theapplicator and out through the circular nozzle orifice [64].The combination of radial and axial forces creates a swirlingthin sheet of liquid that, after some millimeters, becomesunstable and breaks up into “ligaments” before forming theparticles (break-up length). Importantly, a hollow spray coneis formed with particles mainly at the periphery. The keyparameters influencing the properties of the plume andsubsequently the deposition pattern of the particles are theswirl effect, nozzle orifice dimensions, the spray cone angle,and the break-up length. Inthavong et al. [64] reported for aspray with a nozzle diameter of 0.5 mm, a spray cone angleof 30°, and a break-up length of about 3.5 mm, and thediameter at the break-up point is already 4 mm. One studyreported the smallest spray cone diameters (Dmax/Dmin) for aspray angle with 54.6° to be 2.34/1.92 and 3.30/3.08 cm atdistances of 1.0 and 2.5 cm from the nozzle [2]. Anotherstudy reported a spray cone diameter of 2.52/1.58 at 3 cmfrom the nozzle for a spray angle of 39° [65]. Even if thespray pump is inserted as deep as 10–15 mm into the nostril,there is an obvious mismatch between the dimensions andshape of the circular plume (diameter≈2 cm) and the narrowtriangular valve opening. With most of the particles in theperiphery of the plume, it becomes quite evident that themajority of the particles will impinge in the non-ciliatedmucosal walls of the vestibule anterior to the valve. Particlesactually penetrating the valve will do so primarily throughthe lower and wider part of the triangle, a delivery patternthat is accentuated if delivery is performed during sniffing.Although the aerosol-generating mechanisms are different, asimilar mismatch would exist between constricting geome-try of the nasal vestibule and the conical-shaped plumesproduced by other powered devices like pMDIs, nebuliz-ers/atomizers, and many powder devices (see below).

Powered nebulizers and atomizers Nebulizers use com-pressed gasses (air, oxygen, and nitrogen) or ultrasonic ormechanical power to break up medical solutions and

Drug Deliv. and Transl. Res. (2013) 3:42–62 49

suspensions into small aerosol droplets that can be directlyinhaled into the mouth or nose. The smaller particles andslow speed of the nebulized aerosol are advocated to in-crease penetration to the target sites in the middle andsuperior meatuses and the paranasal sinuses [42]. Indeed,nasal inhalation from a nebulizer has been shown to im-prove deposition to the upper narrow part of the nose whencompared to a metered-dose spray pump, but with 33 % and56 % of the delivered dose deposited in the lungs in thesubjects assessed [66]. In light of this problem of lungdelivery, it is unsurprising that nasal inhalation of nebulizedantibiotics intended for topical action in patients with chron-ic rhinosinusitis resulted in coughing and increased need forinhaled medications following nasal inhalation [67].

VibrENT pulsation membrane nebulizer A new nebulizerintended for delivery to the nose and sinuses in patientswith chronic rhinosinusitis utilizing a pulsating aerosol gen-erated via a perforated vibrating membrane has recentlybeen introduced (VibrENT PARI Pharma GmbH). The pul-sation in combination with small particles is assumed tooffer better penetration to the sinuses, and instruction onspecific breathing technique during delivery is advocated tominimize inhalation [68]. Delivery of an aerosol with smallparticles with a mass median aerodynamic diameter(MMAD) of 3.0 μm was performed with two differenttechniques and compared to a spray pump. Aerosol admin-istration into one nostril for 20 s at a rate of mass output of0.3 ml/min, with an exit filter attached to the other nostrilduring nasal breathing, resulted in 4.5 % of the fractiondeposited in the nose (63 %) reaching the sinuses (i.e.,2.8 % of the delivered dose), 27 % in the exit filter, andsignificant lung deposition (10 %). Nasal aerosol deliverywas also performed when the subjects were instructed tomaintain the soft palate closed while a flow resistor wasconnected to the left nostril. Following this procedure, 70 %of the radioactivity was deposited in the nose, 30 % in theexit filter, a negligible fraction in the lungs, and 7 % of thefraction in the nose (i.e., 4.9 % of the delivered dose) wasfound in the sinuses [68]. Following delivery of 100 μl witha traditional spray pump, 100 % of the dose was found in thenose with no deposition in the lungs and non-significantdeposition in the sinuses [68]. Correction for backgroundradiation and decay was performed, but correction for tissueattenuation was not performed, which is likely to change therelative distribution and potentially increase the fraction actu-ally deposited in the lungs [68–71]. Nevertheless, the resultssuggest that the use of a pulsating aerosol in combination withthe breathing technique and an exit resistor may enhancedeposition in the sinuses in healthy volunteers. However, theclinical relevance of these results from healthy volunteers forrhinosinusitis patients with blocked sinus openings remains tobe determined. The proposed breathing technique used to

prevent lung deposition may also prove challenging as com-pared to the automatic integration of velum closure and thedrug delivery process, as achieved when using the exhalationbreath in operation of the delivery device, such as provided byOptiNose’s Bi-Directional™ delivery technology, which canalso utilize an exit resistor to create positive pressure in the noseand sinuses[69]. Furthermore, a very distinct “hot spot” wasobserved for both the nebulizer and spray pump delivery, butno assessment of regional deposition in the nose was performedin the study with the pulsating aerosol nebulizer [68].

Aeroneb Solo vibrating mesh nebulizer Distinct anteriordeposition in the valve area with nebulizers is confirmedin another very recent publication comparing nasal inhala-tion from a nasal sonic/pulsating jet nebulizer (AtomisorNL11S® sonic, DTF-Medical, France) and a new nasal meshnebulizer system designed to minimize lung inhalation(Aeroneb Solo®, Aerogen, Galway, Ireland; DTF-Aerodrug, Tours, France) with the same mean particle size(5.6±0.5 μm) [72]. The new system consists of two inte-grated components: the nebulizer compressor administeringa constant airflow rate transporting the aerosol into onenostril via a nozzle and a pump simultaneously aspiratingfrom a second nozzle in the other nostril at the same airflowrate while the subject is instructed to avoid nasal breathing[72]. The new nasal mesh nebulizer produced more deposi-tion in terms of volume of liquid (27 % vs. 9 %, i.e., 0.81 vs.0.27 ml) in the nasal cavity. The much higher fraction foundin the nasal cavity in this study is probably a result of theshorter nebulizing time and smaller delivered volume in thestudy testing the PARI pulsating nebulizer (20 s at a rate of0.3 ml/min to each nostril versus delivery of 3 ml for up to10 min) before assessment of deposition was performed [68,72]. With much longer delivery time, a substantial fractionof the dose delivered beyond the nasal valve will be clearedto the gastrointestinal (GI) tract.

Aerosol distribution deposition showed a distinct maxi-mum value at 2 cm from the nostril for both nebulizerscorresponding to deposition in the nasal valve region [72].Furthermore, aerosol distribution deposition in the verticalplane showed a similar profile for both nebulizers with adistinct maximum close to the floor of the nose (0.75 cm forthe mesh nebulizer and 1.2 cm for the sonic jet nebulizer)[72]. Importantly, the delivery efficiencies for both nebuliz-ers and delivery techniques appear very low with only 27 %vs. 9 %, i.e., 0.81 vs. 0.27 ml, possibly due to the longdelivery time and resulting differences in mucociliary andother mechanisms of clearance [72]. In other words, a studyassessing deposition after several minutes of delivery islikely to underestimate the actual exposure to the posteriorciliated part of the nose compared to the study assessingdeposition after a short period of delivery of less than 1 min(20 s×2) [68, 72].

50 Drug Deliv. and Transl. Res. (2013) 3:42–62

Clinical relevance of deposition results with nebulizers Lungdeposition and relatively low nasal delivery fractions areissues with nasal nebulizers. Although lung depositionappears to be reduced with simultaneous aspiration fromthe contralateral nostril and with specific breathing instruc-tions, this complex mechanism for use, coupled with theneed for careful patient compliance with breathing, may bechallenging, especially in children or other special popula-tions [66, 68, 72]. The study design, comparing not only twodifferent nebulization techniques but also very differentbreathing techniques, makes interpretation of the resultscomparing the nasal nebulizers in terms of deposition effi-cacy and clinical significance very difficult.

The rationale for using small particles and sonic/pulsa-tion techniques is to increase the delivery into the sinuses,but at the expense of low delivery efficacy and significantpotential for lung deposition. Moreover, despite the intendedadvantages of the vibrating mesh nebulizer that employsaspiration from the contralateral nostril, the quantificationof deposition in the different planes (cartography) demon-strates the typical highly preferential deposition in the ante-rior (anterior 2–3 cm) and lower (lower 1–2 cm) parts of thenasal cavity. This pattern of deposition suggests the nebu-lizer is not effectively delivering to the prime target sites forchronic rhinosinusitis and nasal polyposis (i.e., the middleand superior meatuses or sinuses) [42, 72]. To date, noclinical data has been published with the new nebulizersystems [68, 72].

One approach to avoiding lung deposition is the Bi-DirectionalTM technology employed in OptiNose devices;this technology ensuring operation of the nebulizer only ongeneration of a pressure sufficient to close the palate, avoid-ing the problems associated with suction pumps and specialbreathing instructions. However, clinical data using thisapproach with a nebulizer has also not been published.

ViaNase atomizer A handheld battery-driven atomizerintended for nasal drug delivery has been introduced (Via-Nase by Kurve Technology Inc., Lynnwood, WA, USA).This device atomizes liquids by producing a vortical flow onthe droplets as they exit the device (www.kurvetech.com).The induced vortical flow characteristics can be altered incircular velocity and direction to achieve different droplettrajectories [42, 73]. As discussed above, it is not clear thatvortex flow is desirable for penetration past the nasal valve;however, it has been suggested that this technology is capa-ble of targeting the sinuses, and some gamma-depositionimages suggesting delivery to the sinuses have been pub-lished. However, no information related to impact of priorsurgery or numerical quantification of nasal or sinus depo-sition verifying the claimed improved deposition to theupper parts of the nose has been published [42, 73]. TheViaNase device has been used to deliver nasal insulin in

patients with early Alzheimer’s disease (AD), and clinicalbenefit has been demonstrated [74, 75]. In these studies,delivery of insulin was performed over a 2-min period bynasal inhalation. However, when insulin is delivered withthis device, lung deposition is likely to occur, and someconcerns related to airway irritation and reduction in pul-monary function have been raised in relation to long-termexposure to inhaled insulin when Exubera was marketed fora short period as a treatment for diabetes [71, 76]. Thisexample highlights the issue of unintended lung delivery,one important potential clinical problem associated withusing nebulizers and atomizers producing respirable par-ticles for nasal drug delivery.

Impel nitrogen-driven atomizer A nasal atomizer driven byhighly pressurized nitrogen gas is under development byImpel Inc. (www.impel.com). The device is intended toenable drug delivery to the upper parts of the nose in orderto achieve N2B delivery [77]. To date, only animal data hasbeen presented, making it difficult to evaluate its potential inhuman use, as nasal deposition and the assessment of nasaldeposition in animal models vary significantly fromhumans. As previously noted, however, pMDIs are associ-ated with a number of limitations. It therefore remains to beseen if a pressurized “open-palate” nebulizer will be capableof creating the desired delivery pattern.

Powder devices

Powder medication formulations can offer advantages, includ-ing greater stability than liquid formulations and potential thatpreservatives may not be required. Powders tend to stick to themoist surface of the nasal mucosa before being dissolved andcleared. The use of bioadhesive excipients or agents that slowciliary action may decrease clearance rates and improve ab-sorption [46, 78]. A number of factors like moisture sensitivity,solubility, particle size, particle shape, and flow characteristicswill impact deposition and absorption [3].

The function of nasal powder devices is usually based onone of three principles (Table 1):

1. Powder sprayers with a compressible compartment toprovide a pressure that when released creates a plume ofpowder particles fairly similar to that of a liquid spray;

2. Breath-actuated inhalers where the subject uses his ownbreath to inhale the powder into the nostril from a blisteror capsule; and

3. Nasal insufflators describe devices consisting of amouthpiece and a nosepiece that are fluidly connected.Delivery occurs when the subject exhales into themouthpiece to close the velum, and the airflow carriesthe powder particles into the nose through the devicenosepiece similar to the rhinyle catheter described

Drug Deliv. and Transl. Res. (2013) 3:42–62 51

above. The principle can be applied to different disper-sion technologies and has been further developed andextended into the breath-powered Bi-Directional™ de-livery technology (see below).

Nasal powder inhalers

& Astra Zenaca markets budesonide powder delivered withthe Turbuhaler multi-dose inhaler device modified fornasal inhalation (Rhinocort Turbuhaler®; www.az.com)[79]. It is marketed for allergic rhinitis and nasal polypsin some markets as an alternative to the liquid spray, but itdoes not seem to offer any particular advantage [80]. In astudy comparing twice daily treatment with aqueous bude-sonide spray (128 μg×2) and the Rhinocort Turbuhaler®

(140 μg×2) in nasal polyp patients, both treatments sig-nificantly reduced polyp size compared to placebo, butwith no difference between the active treatments. Howev-er, nasal symptom scores were significantly more reducedin the liquid spray compared to the powder [80]. Agamma-deposition study with Rhinocort Turbuhaler) hasshown predominantly anterior deposition with a “hotspot” at the nasal valve area and about 5% lung deposition[79]. If corrected for tissue attenuation in the lungs, it islikely that the fraction would be substantially higher [69,79].

& Aptar group (www.aptar.com) offers a simple blister-based powder inhaler. The blister is pierced before useand the device nosepiece placed into one nostril. Thesubject closes the other nostril with the finger andinhales the powder into the nose. A powder formulationof apomorphine for Parkinson’s using this blister-basedpowder inhaler (BiDose™/Prohaler™) from Pfeiffer/Aptar was in clinical development by Britannia, a UKcompany recently acquired by Stada Pharmaceutical(www.stada.de). Apparently, further development hasbeen discontinued.

& Nippon Shinyaku Co., Ltd. (www.nippon-shinyaku.-co.jp) markets in Japan a topical steroid (dexamethasonecipecilate) delivered with a powder-based inhalationdevice for allergic rhinitis. The device (Twin-lizer™)has two chambers with capsules inside. The capsule ispierced, and when the subject inhales from the nose-piece, the powder is deagglomerated and delivered intothe nose with the airflow.

Nasal powder sprayers

& SBNL Pharma (www.snbl.com) recently reported dataon a Phase 1 study described in a press release(www.snbl.com) with a zolmitriptan powder cyclodex-trin formulation (μco™ System) for enhanced

absorption, described previously in an in vitro study[81]. The zolmitriptan absorption was rapid, and therelative bioavailability was higher than the marketedtablet and nasal spray (www.snbl.com). The companyhas their own capsule-based, single-dose powder devi-ces (Fit-lizer) [82]. When inserted into a chamber, thetop and bottom of the capsule is cut off by sharp blades.A plastic chamber is compressed by hand, compressedair passes through a one-way valve and the capsuleduring actuation, and the powder is emitted. In vitrotesting shows high-dose reproducibly and minimalresiduals, but no data on particle size distribution or invivo deposition and clearance patterns appear to beavailable. The company has also completed a Phase 2study with the drug granisetron for the indication ofdelayed chemotherapy-induced nausea and vomitingbased on the same formulation technology and deliveredwith the Fit-lizer™ device [81]. They have also an-nounced plans to develop a powder-based influenzavaccine (www.snbl.com).

& Bespak (www.bespak.com), the principle for Unidose-DP™, is similar to the Fit-lizer device. An air-filledcompartment is compressed until a pin ruptures a mem-brane to release the pressure to emit the plume of pow-der. Delivery of powder formulations of a modelantibody (human IgG) has been tested in a nasal castmodel based on human MRI images. Approximately95 % of the dose was delivered to the nasal cavity, butthe majority of it was deposited no further than the nasalvestibule with only about 30 % deposited into deepercompartments of the nasal cavity [83]. The companyreport in their website that they have entered into acollaboration to develop an undisclosed nasal powderproduct with this device (www.bespak.com).

& Aptar group (Pfeiffer/Valois) (www.aptar.com) offers apowder device (Monopowder) based on the same prin-ciple as the devices above but with a plunger that whenpressed creates a positive pressure that ruptures a mem-brane to expel the powder. The device has been used instudies in rabbits, but no data from human deposition orclinical studies have been published [84].

& BD (www.bdpharma.com) also has a powder device(SoluVent™) where a positive pressure is created witha plunger that pierces a membrane to expel the powder.A device based on this technology is being tested withpowder vaccines [85].

Nasal powder insufflators

& Trimel (www.trimel.com) has acquired a device origi-nally developed by a Danish company (Direct Haler).There are two versions of this device that looks like asmall drinking straw. One version is intended for

52 Drug Deliv. and Transl. Res. (2013) 3:42–62

pulmonary drug delivery where subjects inhale throughthe small tubular device and one for nasal drug deliverywhere subjects blow into one end of the tube while theother end is inserted into the vestibule of the nostril. Thedevice can in principle be viewed as a powder version ofthe rhinyle catheter for liquid delivery. This tubulardevice includes a middle section with corrugations.The corrugations allow flexion of the device and createturbulence that deagglomerates the powder. One end ofthe small tubular device is inserted between the lips andthe other into the nasal vestibule. The subject thenexhales through the device to expel the powder fromthe tube and into the nostril. As when using the rhinylecatheter, exhalation into the device causes the soft palateto automatically elevate to separate the oral cavity andthe nasal passages, preventing lung inhalation duringdelivery. No clinical data with the device is available apartfrom a small gamma study in a patent stating that the deviceproduced clearance and areas of deposition that were notsignificantly different from a “state-of-the-art” powder in-halation device (device details not identified) [86].

& OptiNose (www.optinose.com) has developed a breath-powered Bi-Directional™ nasal delivery technology forliquid and powder medications which utilizes the ex-haled breath to deliver the drug into the nose, but withadditional key distinguishing features that importantlyimpact drug deposition and clearance patterns and clin-ical device performance.

Breath-powered Bi-Directional™ technology—a new nasaldrug delivery concept

This novel concept exploits natural functional aspects of theupper airways to offer a delivery method that may overcomemany of the inherent limitations of traditional nasal devices.Importantly, the breath-powered Bi-Directional™ technolo-gy can be adapted to any type of dispersion technology forboth liquids and powders. Breath-powered Bi-Directional™devices consist of a mouthpiece and a sealing nosepiecewith an optimized frusto-conical shape and comfortablesurface that mechanically expands the first part of the nasalvalve (Figs. 1, 2, and 3). The user slides a sealing nosepieceinto one nostril until it forms a seal with the flexible softtissue of the nostril opening, at which point, it mechanicallyexpands the narrow slit-shaped part of the nasal triangularvalve. The user then exhales through an attached mouth-piece. When exhaling into the mouthpiece against the resis-tance of the device, the soft palate (or velum) isautomatically elevated by the positive oropharyngeal pres-sure, isolating the nasal cavity from the rest of the respira-tory system. Owing to the sealing nosepiece, the dynamicpressure that is transferred from the mouth through thedevice to the nose further expands the slit-like nasal

passages. Importantly, the positive pressure in the entrynostril will, due to the sealing nosepiece, balance the oro-pharyngeal pressure across the closed velum to prevent thevelum from being “over-elevated,” thus securing an openflow path between the two nasal passages behind the nasalseptum and in front of the elevated velum.

This “breath-powered” mechanism enables release ofliquid or powder particles into an air stream that enters onenostril, passes entirely around the nasal septum, and exitsthrough the opposite nostril, following a “Bi-Directional™”flow path. Actuation of drug release in devices employingthis approach has been described using manual triggering aswell as mechanisms automatically triggered by flow and/orpressure [13, 69, 70, 87, 88]. By optimizing design param-eters, such as the nosepiece shape, the flow rate, the particlesize profile, and release angle, it is possible to optimizedelivery to target sites beyond the nasal valve, avoid lungdeposition, and to assure that particles are deeply depositedwithout exiting the contralateral nostril. The Bi-Directional™devices currently in phase 3 clinical trials are a multi-doseliquid device incorporating a standard spray pump and acapsule-based powder multi-use device with disposable drugchamber and nosepiece (Fig. 3), but other configurations arepossible. Importantly, the Bi-Directional™ delivery conceptcan be adapted to a variety of dispersion technologies for bothliquids and powders,

Human evidence for nasal deposition patterns with Bi-Directional™ delivery Device variants using this mechanismof nasal drug delivery have been tested in gamma-depositionstudies where assessments of the regional deposition andclearance patterns in human subjects were studied in detail[13, 14, 69]. Comparison of conventional nasal inhalation andBi-Directional™ delivery with the same nebulizer producingsmall particles showed that lung inhalation can be preventedwith Bi-Directional™ delivery even when small respirableparticle are delivered [69]. In one published study, a breath-actuated Bi-Directional™ device incorporating a standardspray pump was compared directly to the same nasal spraypump actuated by hand in the traditional way, and in a secondpublished study, a Bi-Directional™ powder device was di-rectly compared to a traditional spray device [13, 14]. Bothstudies demonstrated less deposition in the non-ciliated nasalvestibule and significantly greater deposition to the upperposterior regions beyond the nasal valve with the Bi-Directional™ devices as compared to conventional deliverywith a spray pump [13, 14] (Fig. 4). In the most recent gammastudy with Bi-Directional™ powder device (Opt-Powder)seen in Fig. 2, the initial deposition in the upper and middleposterior regions of the nose was significantly larger than atraditional spray (upper posterior region; Opt-Powder 18.3±−11.5 % vs. spray 2.4±1.8 %, p<0.02; sum of upper andmiddle posterior regions; Opt-Powder 53.5±18.5 % vs. spray

Drug Deliv. and Transl. Res. (2013) 3:42–62 53

15.7±13.8 %, p<0.02) [14]. In contrast, the summed initialdeposition to the lower anterior and posterior regions for spraywas three times higher compared to Opt-Powder (Opt-Powder17.4±24.5 % vs. spray 59.4±18.2 %, p<0.04; Fig. 4) [14].

Published clinical outcomes with breath-powered Bi-Directional™ delivery devices In addition to human studiesof deposition patterns, devices using the breath-powered Bi-Directional™ technology have also been evaluated in a num-ber of clinical trials. Results generally suggest that superiordeep nasal deposition with clinically important potential canbe achieved in the clinic, and two drug–device combinationsare currently in Phase 3 development: sumatriptan powder for

acute migraine and fluticasone propionate for chronic rhino-sinusitis with nasal polyposis [87–90] (www.optinose.com).

& Midazolam—sedation: Midazolam is a drug with high bio-availability (BA), reasonable ability to cross the BBB, andeasily observed pharmacodynamic effects (sedation). In athree-way crossover study of 12 healthy volunteers, deliveryof the same dose of midazolam (3.4 mg) with a breath-powered Bi-Directional™ device prototype was assessedrelative to a standard nasal spray and intravenous (IV)administration [91]. Drug pharmacokinetics (PK) with bothnasal delivery approaches were similar, as is not unexpectedfor a small molecule easily absorbed to the bloodwith a high

Fig. 3 Cross-sections of a human nose with normal dimensions duringsoft palate closure with Bi-Directional™ flow assessment using CFD.The airflow is entering the right nostril and exiting the left nostril. The

figure illustrates the narrow triangular shape of the nasal valve and thenarrow slit-like passage of the nasal airway more posterior

Fig. 4 Gamma camera image information (logarithmic “hot iron”intensity scale) from the nasal cavity is superimposed on thecorresponding sagittal MRI section. The images are from the samesubject and present deposition 2 min after delivery using (a) a tradi-tional liquid spray, (b) the breath-powered Bi-Directional™ powderdevice, and (c) the breath-powered Bi-Directional™ liquid spray de-vice incorporating the same spray pump as used in a. The initialdeposition following traditional spray was greatest in the lower anterior

regions of the nose, whereas deposition with the Bi-Directional™ deliv-ery devices was greatest in the upper posterior regions of the nose. Theless broad distribution in b following breath-powered Bi-Directional™powder device is believed to be due to the slower clearance for powderthe first 6–8min, reflecting the dissolution of the powder into the mucosallayer. a and b have been published previously, and they are reprinted withpermission from the publisher [14]

54 Drug Deliv. and Transl. Res. (2013) 3:42–62

BA of ≈70 %. Interestingly, the pharmacodynamic effects(onset and level of sedation) reported with Bi-Directional™delivery were very similar to IV administration despitesubstantially lower maximum serum levels (Bi-Direction-al™withmedianCmax03Ng/ml vs. IVwithmedianCmax0

5 ng/ml). In contrast, the onset was slower, and the degree ofsedation was lower following traditional spray deliverydespite similar PK values as Bi-Directional™ delivery[91]. These findings suggest that the sedative effect follow-ing Bi-Directional™ nasal delivery may not merely be aresult of absorption to the blood and subsequent passage intothe brain across the BBB as occurs with a standard nasalspray. Alternative transport routes to the brain bypassing theBBB described in animal studies may contribute to thesedative effects [32–34, 44]. Absorption from the posteriorpart of the nosemay offer amore direct route to brain arterialblood through the particular venous drainage pathway fromthe posterior parts of the nose called “counter-current trans-fer” [32, 33]. Moreover, direct transport to the brain for bothsmall and large molecules may occur along ensheathed cellsforming channels around the olfactory and trigeminal nerves[34, 44]. Contribution from such alternative transport routeswould be consistent with a clinically important improve-ment in the pattern of deep nasal drug deposition withbreath-powered Bi-Directional™ delivery (Fig. 4) [13, 14].

& Sumatriptan—migraine: Unlike midazolam, the seroto-nin antagonist sumatriptan has poor BA when deliveredorally (14 %) and is only marginally higher when deliv-ered as a nasal spray (Pfeiffer single-dose device). It hasbeen estimated that only about 10 % of the drug deliv-ered by standard nasal spray (Imitrex) is absorbed rap-idly across the nasal mucosa within the first 20 min withmuch of a dose undergoing delayed absorption from theGI tract with a Tmax of 90 min [92, 93]. Hypothesizingthat breath-actuated Bi-Directional™ powder deliverymay produce clinically different results than previouslyreported for nasal spray delivery, investigators con-ducted a cross-over PK study in 12 migraineurs, com-paring subcutaneous injection of 6 mg sumatriptan with10 and 20 mg of intranasal sumatriptan powder. Bi-directionally delivered nasal sumatriptan powder waspharmacodynamically similar to injection, inducing asimilar EEG profile and preventing migraine attacks inpatients when delivered 15 min before glyceryl trinitratechallenge. The PK curves showed a similar bi-phasicabsorption pattern as described for sumatriptan nasalspray delivery, but with a substantially higher initialpredominantly nasal absorption peak at 20 min estimat-ed to account for approximately 30 % of the total ab-sorption which is about three times the estimated 10 %fraction absorbed nasally for the marketed Imitrex nasalspray [89, 92]. These PK results lend credence to theconclusion that clinically differentiated nasal deposition

is produced by the breath-powered Bi-Directional™ de-vice compared to what has been previously reportedwith standard nasal spray delivery. A more definitivestudy directly comparing sumatriptan delivery with abreath-powered Bi-Directional™ device to delivery bystandard nasal spray, oral delivery, and injection deliveryis being conducted and should report results soon(www.clinicaltrials.gov). In a randomized, double-blind, parallel group, placebo-controlled study, a singlemigraine attack was treated in-clinic with two doses ofsumatriptan powder (7.5 or 15 mg delivered doses orplacebo) administered intranasally by a novel Bi-Directional™ powder delivery device; fast onset of painrelief was observed for both doses [90]. The pain reliefrates were similar to historical data SC injection despitemuch lower systemic exposure [90, 92]. The resultssuggest that the enhanced deposition associated with thebreath-powered Bi-Directional™ delivery of sumatriptanpowder may contribute to greater initial nasal absorptionand offer clinical benefits [94]. However, based on com-parisons with historical data on the PK and pharmacody-namics profiles of sumatriptan delivered through differentroutes, it has been speculated that the rate of systemicabsorption of nasal sumatriptan may not alone explaindifferences in headache response suggesting the potentialfor an additional route to the site of action as discussedabove [14] . A Phase 3 study is currently in progress(www.clinicaltrials.gov and www.optinose.com).

& Fluticasone propionate—chronic rhinosinusitis with na-sal polyps: Fluticasone is a topical steroid, available as astandard nasal spray for treatment of rhinitis but oftenused with limited benefit in the treatment of chronicrhinosinusitis (CRS) with and without nasal polyps. Ina 3-month placebo controlled study in 109 patients withchronic rhinosinusitis (CRS) with nasal polyps, deliveryof fluticasone (400 μg b.i.d.) with an OptiNose breath-powered Bi-Directional™ liquid drug delivery devicewas reported to be well tolerated and to produce a largemagnitude of reduction in both symptoms and the over-all polyp score. Particularly notable relative to expect-ations with standard nasal spray delivery, completeelimination of the polyps in close to 20 % of the subjectswas reported after 3 months [87]. The proportion ofsubjects with improvement in summed polyp score wassignificantly higher with OptiNose fluticasone propio-nate (Opt-FP) compared with placebo at 4, 8, and12 weeks (22 % vs. 7 %, p00.011, 43 % vs. 7 %, p<0.001, 57 % vs. 9 %, p<0.001). Despite relatively lowerbaseline polyp scores after 12 weeks, the summed polypscore was significantly reduced from 2.8 to 1.8 in the activetreatment group, whereas a minor increase in polyp scorewas seen in the placebo group (−0.98 vs. +0.23, p<0.001).Peak nasal inspiratory flow (PNIF) increased progressively

Drug Deliv. and Transl. Res. (2013) 3:42–62 55

during Opt-FP treatment (p<0.001). Combined symptomscore, nasal blockage, discomfort, rhinitis symptoms, andsense of smell were all significantly improved [87]. Thehighly significant progressive treatment effect of Opt-FPwas observed regardless of baseline polyps score. Previoussinus surgery had no impact on the efficacy. Coupled withthe complete removal of polyps in many patients with smallpolyps, this suggests that improved deposition to target sitesachieved with the Bi-Directional™ delivery device maytranslate into true clinical benefits and possibly reduced needfor surgery [95]. A Phase 3 study is currently in progress(www.clinicaltrials.gov and www.optinose.com).

The same drug–device combination product was alsoevaluated in a small placebo controlled study (N020) inpatients with post-surgical recalcitrant CRS without polyps,producing clinically significant improvements on both ob-jective measures and subjective symptoms [88]. Endoscopyscore for edema showed a significant and progressive im-provement [12 weeks (median scores): Opt-FP −4.0, PBO −1.0, p00.015]. PNIF increased significantly during Opt-FPtreatment compared to placebo (4weeks: p00.006; 8weeks:p00.03). After 12 weeks, MRI scores in the Opt-FP groupimproved against baseline (p00.039), and a non-significanttrend was seen vs. placebo. The nasal RSOM-31 subscalewas significantly improvedwithOpt-FP treatment (4weeks:p00.009, 8 weeks: p00.016, 12 weeks: NS). Sense ofsmell, nasal discomfort, and combined score were all sig-nificantly improved (p<0.05). Notably, this is a conditionmarked by many recent negative placebo-controlled trials[96, 97]. This context, in addition to comparison withhistorical data in similar patient populations, again sug-gests that breath-powered bi-directional delivery is capa-ble of producing superior deep nasal deposition in clinicalpractice (improved targeting of the middle meatus in thiscase) which can translate into improved clinical response(Fig. 4) [13, 87, 88].

& Influenza vaccine: In a four-armed parallel group studywith a whole-virus influenza liquid vaccine without adju-vant, delivery with the breath-powered Bi-Directional™OptiNose device and nasal drops were found to providebetter overall immune response than a traditional nasalspray and an oral spray [50]. In contrast to the self-administration with the OptiNose device, the nasal dropswere delivered by an assistant inserting the pipette tip in acontrolled manner beyond the nasal valve with the neckextended. These results suggest that Bi-Directional™devices are a practical delivery method capable of pro-ducing a clinically relevant broader and deeper distribu-tion of vaccines to the nasal respiratory mucosa, areas richin dendritic cells and aggregates of lymphoid tissue, of-fering potential for a range of vaccines to produce im-proved immune response in non-parenteral delivery forms[24, 50].

Assessment of nasal deposition and clearance—clinicalaspects

CFD simulations

With development of high-resolution CT and MRI technolo-gy, it has become possible to generate accurate 3D reconstruc-tions of the complex nasal anatomy (Fig. 3). The field ofcomputational fluid dynamics (CFD) is rapidly progressingin medicine and has enabled CFD simulations of nasal aero-dynamics and deposition patterns [98–101]. The greatly im-proved density of the grids used and algorithms, along withmuch faster computers available for simulation, now allowimplementation of more realistic conditions. For example,recent publications describe algorithms to simulate septalabnormalities, post-surgical changes, as well as heat and waterexchange, and to more accurately simulate the true propertiesof aerosol generation and plume characteristics [99–101].Undoubtedly, as the quality and capabilities increase, CFDsimulations will play an increasingly important role and allowfor realistic simulation of nasal physiology and drug delivery.A more detailed review of this exciting field is outside thescope of this review.

Deposition studies in casts

The progress in imaging and reconstruction software hasalso made it possible to make physical models in rigidmaterials by modern 3D printing techniques like stereoli-thography with correct nasal geometry and dimensions.Casts made in softer material like silicone may offer advan-tages in terms of more realistic device cast interface. How-ever, caution is necessary because even the softer siliconecasts do not realistically represent the nasal valve dynamics,the cyclic physiological changes of the mucosa, or reflectthe in vivo surface properties of the nasal mucosa, includingthe impact on mucocliliary clearance [102].

An in depth review of in vitro drug delivery simulationperformed in nasal casts is also outside the scope of thisreview, but some comments related to recent work areincluded to highlight issues related to the interpretationand predictive value of results obtained with nasal deliverydevices in cast studies. Three recent publications report indetail on the effect of breathing patterns, formulation, spraypump variables, and the site of deposition in a particularcommercially available silicone cast (Koken Co., Japan)[65, 103, 104]. An interesting gel coating method thatchanges color in contact with the liquid allowing quantifi-cation of deposition by photometric analysis of depositionimages is described [103]. In the most recent work, differentinsertion depth, spray angle, and plume characteristics (coneangle and particle size distribution) were studied. Data onthe dimensions of the cast are not presented in these reports;

56 Drug Deliv. and Transl. Res. (2013) 3:42–62

however, it is critical to note that the Koken cast is, accord-ing to the manufacturer, primarily an educational tool andthat it therefore has a flat transparent septum to enablevisualization of complicated nasal structures. Inspection ofthe nasal valve area and objective measurements of thedimensions reveals that the dimensions at the valve areaare several-fold larger than the average human valve dimen-sions and outside the normal range [105]. It is suggested inthese recent publications that casts studies have potential forestablishing in vivo bioequivalence and as indicators ofcritical quality attributes [65]. While an admirable goal,the lack of validation of all cast dimensions coupled withthe inability of the cast to reproduce important dynamicaspects of nasal anatomy and physiology discussed previ-ously, certainly casts doubt on the ability to achieve thisobjective with the Koken cast, and potentially any rigidnasal cast. Nevertheless, the use of ever-improving castscoupled with innovative techniques such as photometricsmay be very useful in development of new nasal deliverydevices. Reliance on standards published by FDA for per-formance of spray pumps may seem appropriate for com-parison of nasal delivery devices; however, publishedanalysis also suggests that the in vitro measurements in theFDA guidance related to performance of spray pumps arenot clinically relevant [2]. Thus, in light of current method-ological and technological limitations, human in vivo depo-sition and clearance studies, and relevant human clinicaltrials, allowing regional deposition quantification and directclinical comparisons, respectively, are still ultimately re-quired. A recent review concludes that although both invitro studies and in vivo imaging methods may be of valueduring the device development stages, ultimately, random-ized placebo-controlled trials quantifying both symptomsand functional parameters are required to determine drugdelivery efficiency of different devices [42].

In vivo assessment of deposition and clearance

A number of gamma deposition studies, a study usingradiopaque contrast, and studies using colored dyes confirmthat administration with conventional spray pumps, pMDIs,nebulizers, and powder devices all result in deposition main-ly in the anterior non-ciliated segments of the nose anteriorto and at the narrow nasal valve, which is regarded subop-timal for clinical efficacy where deep and broad nasal de-position is required [13, 43, 63, 66, 72, 79, 106]. Coloreddyes may offer a quick and inexpensive semi-quantitativeassessment of deposition and clearance, and a number ofstudies have assessed deposition patterns with dyes with thegoal of improving deposition and the clinical outcome ofdelivery with spray pumps and drops [43, 107, 108]. Al-though results vary, the effect of different body positionsand administration techniques appears to have limited

impact on initial deposition patterns. In fact, a recentsingle-blind, cross-over study comparing seven differentadministration techniques of colored dyes in healthy indi-viduals using endoscopic video imaging concluded thatthere may not be a single “best” technique for topical nasaldrug delivery with conventional nasal sprays [108]. Lack ofpatient compliance further reduces the clinical usefulness ofthese delivery techniques.

More detailed assessment of drug deposition using regionalgamma-deposition patterns have added to the understandingof deposition and clearance patterns and how they may havean impact on the clinical outcomes [13, 14, 66, 70, 72].Improved methods for positioning and re-positioning of thetest subjects and the use of radiolabeled gases and MRIoverlay allow regional quantification of nasal deposition andoutcomes [66, 70]. Furthermore, in contrast to earlier studies,proper correction for regional differences in tissue attenuationin the different nasal segments and between the nose and lungsis now being performed [13, 14, 70]. This review onlyaddresses in vivo gamma-deposition studies dealing withsome key aspects related to the in vivo performance of nasaldelivery devices that normally get limited attention.

Impact of delivery instructions, patient compliance,and body position

One factor too often neglected when comparing depositionstudies is whether the delivery procedure was performed bythe subjects themselves or by an assistant. Clearly, deliveryby the subjects is much closer to the real-life situation, butinevitably introduces more variability. In most gamma-deposition studies, a trained assistant inserts the spray de-vice and performs the actuation according to a strict proto-col. This was the case in a study assessing deposition ofradiolabeled cromoglycate substantial delivery beyond thenasal valve along the nasal floor was observed [109]. Incontrast, in a study with radiolabeled insulin where the spraywas actuated by the subjects themselves, it was noted thatindividual administration technique resulted in the majorityof doses being deposited in the anterior rather than theposterior nasal cavity in five out of six subjects, with thedose then being cleared via the nares rather than the naso-pharynx [110]. Contrary to expectations, no sign of systemicabsorption of insulin was observed, and the authors com-mented that this effect of individual administration tech-nique raises a separate question on the usefulness of nasalspray doses for delivery of insulin intended for systemicabsorption [110].

Overall versus regional clearance patterns

Gamma studies must be performed in a controlled settingwhere subjects are more likely to adhere to instructions for

Drug Deliv. and Transl. Res. (2013) 3:42–62 57

use of the devices than in real life. It is very common toobserve that subjects during, or immediately after, adminis-tration of drug using nasal devices intuitively sniff to avoidthe concentrated anterior liquid deposition from drippingout and down on the upper lips. Sometimes, the anteriorlydeposited surplus is wiped off, as has been observed ingamma-deposition studies [111]. In fact, considerable earlydrip-out has been observed in a gamma study following self-administration with a 100-μl standard nasal spray pump,which causes concentrated anterior deposition. This phe-nomenon has also been observed after delivery with nebu-lizers [14, 72]. Recent studies offering regional clearancecurves for four or six nasal segments highlight that the initialsite of deposition has a major impact on the clearance ratesand that determination of overall nasal clearance is a verycrude and potentially misleading measure that does notpredict clinical performance [13, 14]. Interestingly, a recentreview on pulmonary drug delivery states that total lungdeposition appears to be a poor predictor of clinical out-come; rather, regional deposition needs to be assessed topredict therapeutic effectiveness [112]. In a study comparingnasal deposition and clearance after self-administration ofthe same conventional spray pump (100 μl) by hand in thetraditional way and by breath actuation with a Bi-Directional™ delivery device (see below for details), thepercentage left in the nose 30 min after hand actuation istwice that of breath actuation (46 % vs. 23 %). However, theregional deposition patterns (divided in four nasal segments)reveal that this difference is primarily a result of anteriorretention in the predominantly non-ciliated anterior twonasal quadrants following hand-actuated spray delivery.The deposit ion pattern is reversed with the Bi-Directional™ device, which was reported to offer threetimes greater broader and more reproducible deposition tothe ciliated respiratory mucosa beyond the nasal valve and,in particular, in the upper posterior segments, with removalat a speed corresponding to expected mucocliliary clearancerate [13]. Another study comparing self-administration of aspray pump and a Bi-Directional™ breath-actuated powderdevice showed a similar significant difference in the region-al deposition and clearance patterns, further reinforcing theimportance of evaluating not only overall or “whole-nose”deposition and clearance but instead also evaluating region-al patterns when developing or comparing nasal deliverydevices [14] (Fig. 4).

Impact of site of delivery and volume on depositionand clearance

The results from the study described above comparing de-position and clearance after delivery from the same spraypump actuated in different manners show that the initial siteof deposition has a profound impact on the clearance rates

[3, 13, 14]. Interestingly, McLean et al. [113] describedthree different phases of nasal clearance.

1. The first phase occurs within the first minute afteradministration and is particularly evident following de-livery of large concentrated volumes that rapidly passalong the floor of the nose to the pharynx to be swal-lowed. This applies in particular to delivery of dropsand can contribute to explaining the much lower ab-sorption of desmopressin delivered as drops, but alsoapplies to spray delivery with higher spray volumes [3,14, 51, 113]. The initial and very rapid removal may notalways be recognized, as the initial gamma image oftenincludes averaging of registration of counts over a 2-min period due to the relatively small dose of radioac-tivity used (for ethical reasons) [14].

2. The second phase lasts for about 15 min and correspondsto mucociliary clearance of the fraction initially depositedon the ciliated respiratory mucosa found at and beyondthe nasal valve [3, 13, 14, 51, 63, 70, 113, 114].

3. The third prolonged late phase represents the slow re-moval of residual drug deposited in the anterior non-ciliated parts of the nasal surface and can take hours,unless mechanically removed by nose blowing and/orwiping of the nose [63]. Consequently, depending onwhether the substance in question has local action, isintended for systemic absorption, for N2B transport, ora combination, the primary goal is frequently to maxi-mize exposure to the ciliated mucosa beyond thenasal valve. One strategy for enhanced exposure isto slow clearance by thixotropic or bioadhesiveagents or agents which slow ciliary action in orderto increase the residence time in this region or byadding absorption enhancer if systemic absorption isthe objective [78, 115].

In principle, an alternative, complementary, and probablybetter way to enhance the exposure is to modify/improve theadministration method or technique. The goal should be toreduce the amount of drug quickly passing through the noseto be swallowed in the first phase, to reduce the amountdeposited outside the nose, and to increase the amountbypassing the nasal valve and the nasal respiratory mucosalsurface covered. Delivery of smaller particles with a tradi-tional spray offers advantages in terms of absorption andbiological response compared to delivery of drops, and re-peated delivery of a smaller volume, as 2×50-μl spray hasbeen reported to be better than 1×100 μl for systemic absorp-tion [51, 114]. In contrast, another study found that spraying1×100 μl resulted in larger deposition than 2×50 μl beyondthe nasal valve with more rapid overall clearance, but thestudy did not assess absorption or biological response [63].A narrow cone angle resulted in more posterior deposition and

58 Drug Deliv. and Transl. Res. (2013) 3:42–62

faster clearance than a cone of 60°, and drops deposited moreposteriorly are cleared faster [116, 117] .

For locally acting anti-inflammatory drugs like steroidsand antihistamines, as well as for vaccines, the non-ciliatedsurface of the vestibule is not the target [42]. However,recent publications continue to advocate concentrated ante-rior deposition and retention as desirable and a key advan-tage of the novel HFA-based nasal pMDI with topicallyacting drug [118]. Reference is made to a paper from 1987with CFC-based pMDI showing that as much as 65 % of theinitial radioactivity is retained in the anterior parts of thenose after 30 min and incorrectly stating that an almost totalclearance was observed 30 min after delivery with aqueousspray [63]. A recent publication even claims that the anteriorretention following pMDI delivery provides evidence forenhanced efficacy, which seems to be in conflict with theprevailing opinion [42, 118].

Conclusions

The nose is attractive for delivery of many drugs andvaccines, but the potential has not been fully realized.Inherent challenges related to the nasal anatomy, physiol-ogy, and aerodynamics that may severely limit the poten-tial and clinical efficiency are not widely understood. Thesmall and dynamic dimensions of the nasal cavity and theanterior anatomy are among the most important hurdlesfor more efficient nasal drug delivery. Despite importantimprovements in the technical device attributes that canoffer more reproducible and reliable in vitro performance,this has to a limited extent translated into improved clin-ical performance. While in vitro performance testing isundoubtedly of value for product quality assessment, pre-dictive value for in vivo clinical performance is highlyquestionable [2]. CFD simulations of nasal aerodynamicsand cast studies may be of value in the developmentalstages of device design, and future advances may improvetheir predictive value. Human in vivo deposition andclearance studies can be very important, providing valu-able information particularly if recent advances allowingregional quantification and tissue attenuation correctionare employed [14, 70, 112]. Still, delivery by trainedassistants in controlled environments may not adequatelyreflect the device performance in the clinical setting. Eventhe most advanced nebulizer technologies introduced haveshown poor delivery efficiency, with undesirable localizeddelivery in the non-ciliated anterior nasal region and alongthe floor of the nose and problems with inhalation expo-sure of the lungs [72]. As stated in a recent review, well-controlled clinical studies are currently required to quan-tify changes in both symptoms and functional parameters,and ultimately to determine the efficacy of novel drug/

device combinations [42]. The Bi-Directional™ drug de-livery concept introduces a novel approach that can over-come inherent limitations of conventional nasal deliveryimposed by the dynamics of the nasal valve. Gamma-scintigraphy studies with both powder and liquid Bi-Directional™ device variants confirm significant improve-ments in regional in vivo deposition and clearance pat-terns, and a number of clinical trials suggest that this deepnasal deposition translates into clinical benefits for thepatients. This delivery technology can be combined witha variety of dispersion technologies for both liquids andpowders, and promises to expand the possibilities of nasaldrug delivery.

Conflict of interest P.G. Djupesland is a founder, CSO and share-holder of OptiNose, a commercial company developing nasal deliverydevices.

Open Access This article is distributed under the terms of the Crea-tive Commons Attribution License which permits any use, distribution,and reproduction in any medium, provided the original author(s) andthe source are credited.

References

1. FDA 2003. US FDA draft guidance for industry. Bioavailabilityand bioequivalence studies for nasal aerosols and nasal sprays.Bethesda. http://www.fda.gov/cder/guidance/index.htm.Accessed July 2012.

2. Suman JD, Laube BL, Dalby R. Validity of in vitro tests onaqueous spray pumps as surrogates for nasal deposition, absorp-tion and biologic response. J Aerosol Med. 2006;19:510–21.

3. Vidgren MT, Kublik H. Nasal delivery systems and their effect ondeposition and absorption. Adv Drug Deliv Rev. 1998;29:157–77.

4. Sahin-Yilmaz A, Naclerio RM. Anatomy and physiology of theupper airway. Proc Am Thorac Soc. 2011;8:31–9.

5. Cole P. Nasal respiratory function. In: The nose. St. Louis:Mosby-Year Book Inc.; 1993. p. 3–60.

6. Cole P. The mouth and the throat. In: The nose. St. Louis: Mosby-Year Book Inc.; 1993. p. 61–90.

7. Haight JS, Cole P. The site and function of the nasal valve.Laryngoscope. 1983;93:49–55.

8. Yu S, Liu Y, Sun X, Li S. Influence of nasal structure on thedistribution of airflow in nasal cavity. Rhinology. 2008;46:137–44.

9. Cole P. The four components of the nasal valve. Am J Rhinol.2003;17:107–10.

10. Fodil R, Brugel-Ribere L, Croce C, Sbirlea-Apiou G, Larger C,Papon JC, Delclaux C, Coste A, Isabay D, Louis B. Inspiratoryflow in the nose: a model coupling flow and vasoerectile tissuedistensibility. J Appl Physiol. 2005;98:288–95.

11. Cole P. Nasal and oral airflow resistors. Site, function, and as-sessment. Arch Otolaryngol Head Neck Surg. 1992;118:790–3.

12. Gomes AC, Sampaio-Teixeira ACM, Trindade SHK, TrindadeIEK. Nasal cavity geometry of healthy adults assessed using acous-tic rhinometry. Rev Bras Otorrinolaringol. 2008;74(5):746–54.

13. Djupesland PG, Skretting A, Windern M, Holand T. Breathactuated device improves delivery to target sites beyond the nasalvalve. Laryngoscope. 2006;116(3):466–72.

14. Djupesland PG, Skretting A. Nasal deposition and clearance inman: comparison of a bidirectional powder device and a traditional

Drug Deliv. and Transl. Res. (2013) 3:42–62 59

liquid spray pump. J Aerosol Med Pulm. 2012. doi:10.1089/jamp.2011.0924.

15. Swift DL, Proctor DF. Access of air to the respiratory tract. In:Brain DJ, Proctor DF, Reid LM, editors. Respiratory defensemechanisms. New York: Marcel Dekker; 1977. p. 63–93.

16. Mann DG, Sasaki CT, Fukuda H, Suzuki M, Hernandez JR.Dilator naris muscle. Ann Otol Rhinol Laryngol. 1977;86:362–70.

17. Djupesland PG, Skatvedt O, Brogersen AK. Dichotomous phys-iological effects of nocturnal external nasal dilation in heavysnorers: the answer to a rhinological controversy? Am J Rhninol.2001;15(2):95–103.

18. Hairfield WM, Warren DW, Hinton VA, Seaton DL. Inspiratoryand expiratory effects of nasal breathing. Cleft Palate J.1987;24:183–9.

19. Mygind N, Dahl R. Anatomy, physiology and function of thenasal cavities in health and disease. Adv Drug Deliv Rev.1998;29:3–12.

20. Proctor DF. The mucociliary system. In: Proctor DF, Andersen I,editors. The nose: upper airway physiology and the atmo-spheric environment. Amsterdam: Elsevier Biomedical; 1982.p. 245–78.

21. Halama AR, Decreton S, Bijloos JM, Clement PAR. Density ofepithelial cells in the normal human nose. A scanning electronmicroscopic study. Rhinology. 1990;28:25–32.

22. Svensson S, Olin AC, Hellgren J. Increased net water loss by oralcompared to nasal expiration in healthy subjects. Rhinology.2006;44:74–7.

23. Andersen I, Lundqvist GR, Jensen PL, Proctor DF. Human re-sponse to controlled levels of sulfur dioxide. Arch Environ Hlth.1974;28:31–9.

24. Brandtzaeg P. Role of secretory antibodies in the defense againstinfections. Int J Med Microbiol. 2003;293:1–13.

25. Baraniuk JN. Neural regulation of mucosal function. Pulm Phar-macol Ther. 2008;21(3):442–8.

26. Cole P. Stability of nasal airflow. Clnin Otolaryngol. 1989;14:177–82.27. Cole P, Haight JSJ. Posture and the nasal cycle. Ann Otol Rhinol

Laryngol. 1986;95:223–7.28. Havas TE, Cole P, Gullane PJ, et al. The nasal cycle after

laryngoectomy. Acta Otolaryngo (Stockh). 1987;103:111–6.29. Djupesland PG, Chatkin JM, Qian W, Haight JSJ. Nitric oxide in

the nasal airway: a new dimension in otolaryngology. Am JOtolaryngol. 2001;22:19–32.

30. Qian W, Sabo R, Ohm M, Haight JSJ, Fenton R. Nasal nitricoxide and the nasal cycle. Laryngoscope. 2001;111:1603–7.

31. Jones N. The nose and paranasal sinuses physiology and anato-my. Adv Drug Deliv Rev. 2001;51:5–19.

32. Einer-Jensen N, Larsen L. Local transfer of diazepam, but not ofcocaine, from the nasal cavities to the brain arterial blood in rats.Pharmacol Toxicol. 2000;87:276–8.

33. Einer-Jensen N, Larsen L, Deprez S, Starns E, Schwartz S.Intranasal absorption of sumatriptan and naratriptan: no evidenceof local transfer from the nasal cavities to the brain arterial bloodin male rats. Biopharm Drug Dispos. 2001;22(5):213–9.

34. Dhuria S, Hanson LR, Frey WH. Intranasal delivery to the centralnervous system: mechanisms and experimental considerations. JPharm Sci. 2009. doi:10.1002/jps.21924.

35. Landis BN, Scheibe M, Weber C, Berger R, Brämerson A, BendeM, Nordin S, Hummel T. Chemosensory interaction: acquiredolfactory impairment is associated with decreased taste function.J Neurol. 2010;257:1303–8.

36. Brant JG. Within reach of an end to unnecessary bitterness.Lancet. 2000;356:1371–2.

37. Feron F, Perry C, McGrath JJ, Mackay-Sim A. New techniquesfor biopsy and culture of human olfactory epithelial neurons.Arch Otolaryngol Head Neck Surg. 1988;124(8):861–6.

38. Leopold DA, Hummel T, Schwob JE. Anterior distribution ofhuman olfactory epithelium. Laryngoscope. 2000;110:417–21.

39. Hummel T, Livermore A. Intranasal chemosensory function ofthe trigeminal nerve and aspects of its relation to olfaction. IntArch Occup Environ Health. 2002;75:305–13.

40. Husner A, Frasnelli J, Welge-Lussen A, Reiss G, Zahnert T,Hummel T. Loss of trigeminal sensitivity reduces olfactory func-tion. Laryngoscope. 2006;116:1520–2.

41. Waddell AN, Patel SK, Toma AG, Maw AR. Intranasal steroidspary in the treatment of rhinitis: is one better than another? JLaryngol Otol. 2003;117:8843–5.

42. Laube B. Devices for aerosol delivery to treat sinusitis. J AerosolMed. 2007;20(Suppl):5–18.

43. Aggrawal R, Cardozo A, Homer JJ. The assessment of topicalnasal drug distribution. Clin Otolaryngol. 2004;29:201–5.

44. Johnson HJ, Hanson LR, Frey WH. Trigeminal pathways delivera low molecular weight drug from the nose to the brain andorofacial structures. Mol Pharm. 2010;7(3):884–93.

45. Marple B, Roland P, Benninger M. Safety review of benzalko-nium chloride used as a preservative in intranasal solutions: anoverview of conflicting data and opinions. Otolaryngol HeadNeck Surg. 2004;130:131–41.

46. Illum L. Nasal drug delivery: possibilities, problems and solu-tions. J Control Release. 2003;87:187–98.

47. Penttilä M, Poulsen P, Hollingworth K, Holmström M. Dose-related efficacy and tolerability of fluticasone propionate nasaldrops 400 μg once daily and twice daily in the treatment ofbilateral nasal polyposis: a placebo-controlled randomized studyin adult patients. Clin Exp Allergy. 2000;30:94–102.

48. Keith P, Nieminen J, Hollingworth K, Dolovich J. Efficacy andtolerability of fluticasone propionate nasal drops 400 μg oncedaily compared with placebo for the treatment of bilateral poly-posis in adults. Clin Exp Allergy. 2000;30:1460–8.

49. Merkus P, Ebbens FA, Muller B, Fokkens WJ. Influence ofanatomy and head position on intranasal drug deposition. EurArch Otorhinolaryngol. 2006;263:827–32.

50. Bakke B, Samdal HH, Holst J, et al. Vaccine oral spray immuni-zation may be an alternative to intranasal vaccine delivery toinduce systemic antibodies but not nasal mucosal or cellularimmunity. Scan J of Immunol. 2006;63:223–31.

51. Harris AS, Nilsson IM, Wagner ZG, Alkner U. Intranasaladministration of peptides: nasal deposition, biological re-sponse and absorption of desmopressin. J Pharm Sci. 1986;75:1085–8.

52. Berger WE, Godfrey JW, Slater AL. Intranasal corticosteroids:the development of a drug delivery device for fluticasone furoateas a potential step toward improved compliance. Expert OpinDrug Deliv. 2007;4(6):689–701.

53. Kanowitz SJ, Batra PS, Citardi MJ. Topical budesonide viamucosal atomization device in refractory postoperativechronic rhinosinusitis. Otolaryngol Head Neck Surg. 2008;139:131–6.

54. Renteria SS, Clemens CC, Croyle MA. ∗Development of a nasaladenovirus-based vaccine: effect of concentration and formula-tion on adenovirus stability and infectious titer during actuationfrom two delivery devices. Vaccine. 2010;28:2137–48.

55. Nichol KL, Mendelman PM, Mallon KP, et al. Effectiveness oflive, attenuated intranasal influenza virus vaccine in healthy,working adults: a randomized controlled trial. JAMA. 1999;282:137–44.

56. Belshe RB, Edwards KM, Vesikari T, et al. Live attenuated versusinactivated influenza vaccine in infants and young children. NEngl J Med. 2007;356:685–96.

57. Mutsch M, Zhou W, Rhodes P, et al. Use of the inactivatedintranasal influenza vaccine and the risk of Bell’s palsy in Swit-zerland. N Engl J Med. 2004;350:896–903.

60 Drug Deliv. and Transl. Res. (2013) 3:42–62

58. Wermeling DP, Miller JL, Archer SM, Rayens MK, Rudy AC.Pharmacokinetics, bioequivalence, and spray weight reproduc-ibility of intranasal butorphanol after administration with 2 dif-ferent nasal spray pumps. J Clin Pharmacol. 2005;45:969–73.

59. Rapoport A, Winner P. Nasal delivery of antimigraine drugs:clinical rationale and evidence base. Headache. 2006;46 Suppl4:S192–201.

60. Hankin CS, Cox L, Lang D, Bronstone, Wang Z, Lepore MS,Buck PO. Medical costs and adherence in patients receivingaqueous versus pressurized aerosol formulations of intranasalcorticosteroids. All Asthma Proc. 2012;33:258–64.

61. Mygind N, Vesterhauge S. Aerosol distribution in the nose.Rhinology. 1978;16:79–88.

62. Meltzer EO, Jacobs RL, LaForce CF, Kelley CL, Dunbar SA,Tantry SK. Safety and efficacy of once-daily treatment withbeclomethasone dipropionate nasal aerosol in subjects with pe-rennial allergic rhinitis. Asthma Proc. 2012;33:249–57.

63. Newman SP, Moren F, Clarke SW. The nasal distribution ofmetered dose inhalers. J Laryngol Otol. 1987;101:127–32.

64. Inthavong K, Tian ZF, Tu JY, Yang W, Xue C. Optimising nasalspray parameters for efficient drug delivery using computationalfluid dynamics. Computers in Biol Med. 2008;38:713–26.

65. Kundoor V, Dalby RN. Effect of formulation- and administration-related variables on deposition pattern of nasal spray pumpsevaluated using a nasal cast. Pharm Res. 2011;28:1895–904.

66. Suman J, Laube BL, Dalby R. Comparisonof nasal depositionand clearance of aerosol generated by nebulizer and aqueousspray pump. Pharm Res. 1999;16:1648–52.

67. Vaughan WC, Carvalho G. Use of nebulized antibiotics for acuteinfections in chronic sinusitis. Otolaryngol Head Neck Surg.2002;127:558–68.

68. MöllerW, SabaGK, Häussinger K, Becker S, KellerM, SchuschnigU. Nasally inhaled pulsating aerosols: lung, sinus and nose deposi-tion. Rhinology. 2011;49:286–91.

69. Djupesland PG, Skretting A, Windern M, Holand T. A novelconcept for nasal delivery of aerosols can prevent lung inhalation.J Aerosol Med. 2004;17(3):249–59.

70. Skretting A, Djupesland PG. A new method for scintigraphicquantification of deposition and clearance in anatomical regionsof the human nose. Nucl Med Comm. 2009;30(8):629–38.

71. Djupesland PG, Craft S. Intranasal insulin improves cognitionand modulates β-amyloid in early AD. Letters to the editor.Neurology. 2008;71:864.

72. Vecellio L, De Gersem R, Le Guellec S, et al. Deposition ofaerosols delivered by nasal route with jet and mesh nebulizers.Int J Pharm. 2011;407:87–94.

73. Giroux M. Controlled particle dispersion: effective nasal deliveryfrom a versatile, flexible delivery platform. OndrugDelivery.www.ondrugdelivery.com. 2005;13–15.

74. Reger MA, Watson GS, Green PS, et al. Intranasal insulinimproves cognition and modulates beta-amyloid in early AD.Neurology. 2008;70(6):440–8.

75. Craft S, Baker LD, Montine TJ, et al. Intranasal insulin therapyfor Alzheimer disease and amnestic mild cognitive impairment: apilot clinical trial. Arch Neurol. 2012;69(1):29–38.

76. Selam J. Inhaled insulin: promises and concerns. J Diabetes SciTechnol. 2008;2(2):311–5.

77. Hoekman JD, Ho RJY. Enhanced analgesic responses after pref-erential delivery of morphine and fentanyl to the olfactory epi-thelium in rats. Anesth Analg. 2011;113:641–51.

78. Soane RJ, Frier M, Perkins AC, Jones NS, Davis SS, Illum L.Evaluation of the clearance characteristics of bioadhesive systemsin humans. Int J Pharma. 1999;178:55–65.

79. Thorsson L, Newman SP, Weisz A, Trofast E, Moren F. Nasaldistribution of budesonide inhaled via a powder inhaler. Rhinol-ogy. 1993;31:7–10.

80. Tos M, Svendstrup F, Arndal H, et al. Efficacy of an aqueous anda powder formulation of nasal budesonide compared in patientswith nasal polyps. Am J Rhinol. 1998;12:183–9.

81. Cho HJ, Balakrishnan P, Shim WS, Chung SJ, Shim CK, KimDD. Characterization and in vitro evaluation of freeze-driedmicroparticles composed of granisetron-cyclodextrin complexand carboxymethylcellulose for intranasal delivery. Int J Pharm.2010;400(1–2):59–65.

82. Haruta S, Tsutsui T. Meeting the needs for nasal delivery devicesfor powder formulations. Drug Dev Deliv. 2012;12(5):22–7.

83. Kaye RS, Purewal TS, Alpar OH. Development and testing ofparticulate formulations for the nasal delivery of antibodies. JControl Rel. 2009;135:127–35.

84. Pringels E, Callens C, Vervaet C, Dumont F, Slegers G, ForemanP, Remona JP. Influence of deposition and spray pattern of nasalpowders on insulin bioavailability. Int J Pharma. 2006;310:1–7.

85. Huanga J, Garmise RJ, Crowder TM, Mara K, Hwanga CR,Hickey AJ, Mikszta JA, Sullivana VJ. A novel dry powderinfluenza vaccine and intranasal delivery technology: inductionof systemic and mucosal immune responses in rats. Vaccine.2004;23:794–801.

86. Keldmann E. Patent application. Inhaler for powdered medica-ment. WO 98/53869. 1998

87. Vlckova I, Navrátil P, Kaňa R, et al. Effective treatment of mild-to-moderate nasal polyposis with fluticasone delivered by a noveldevice. Rhinology. 2009;47:419–26.

88. Hansen F, Djupesland PG, Fokkens WJ. Effective treatment ofchronic rhinosinusitis with fluticasone delivered by a noveldevice: a randomized placebo controlled pilot study. Rhinology.2010;48:292–9.

89. Luthringer R, Djupesland PG, Sheldrake CD, et al. Rapidabsorption of sumatriptan powder and effects on glyceryltrini-trate model of headache following intranasal delivery using anovel bi-directional device. J Pharm Pharmacol. 2009;61:1219–28.

90. Djupesland PG, Dočekal P. Intranasal sumatriptan powder deliv-ered by a novel breath actuated Bi-Directional device for theacute treatment of migraine: a randomised, placebo-controlledstudy. Cephalalgia. 2010. doi:10.1177/0333102409359314.

91. Dale O, Nilsen T, Loftsson T, Tønnesen HH, et al. Intranasalmidazolam: a comparison of two delivery devices in humanvolunteers. J Pharma Pharmacol. 2006;58:1311–8.

92. Fox AW. Onset of effect of 5-HT1B/1D agonists: a model withpharmacokinetic validation. Headache. 2004;44:142–7.

93. Duquesnoy C, Mamet JP, Sumner D, Fuseau E. Comparativeclinical pharmacokinetics of single doses of, sumatriptan follow-ing subcutaneous, oral, rectal and intranasal administration. Eur JPharm Sci. 1998;6:99–104.

94. Djupesland PG, Mahmoud R, Messina J, Miller P. Rate of sys-temic absorption of sumatriptan may not explain differences inheadache response suggesting the potential for an additional routeto the site of action. Cephalalgia. 2011;31 Suppl 1:PS1–PS55.

95. Djupesland PG, Vlckova I, Hewson G. Impact of baseline nasalpolyp size and previous surgery on efficacy of fluticasone deliv-ered with a novel device: a subgroup analysis. A J Rhinol Aller-gy. 2010. doi:10.2500/ajra.2010.24.3516.

96. Fokkens WJ. Unmet needs in rhinology. Rhinology. 2008;46:257–8.97. Fokkens WJ. Recalcitrant rhinosinusitis, the diagnosis and treat-

ment and evaluation of results. Rhinology. 2010;48:257–8.98. Kleven M, Djupesland PG, Melaaen M. Computational fluid dy-

namics (CFD) applied on the drug delivery design process to thenasal passages: a review. J Mech Med Biol. 2012. doi:10.1142/S0219519411004526.

99. Frank DO, Kimbell JS, Pawar S, Rhee JS. Effects of anatomy andparticle size on nasal sprays and nebulizers. Oto Head Neck Surg.2012;146(2):313–9.

Drug Deliv. and Transl. Res. (2013) 3:42–62 61

100. Inthavong K, Ge Q, Se CMK, Yang W, Tu JY. Simulation ofsprayed particle deposition in a human nasal cavity including anasal spray device. J Aerosol Sci. 2011;42:100–13.

101. Sommer F, Kröger R, Lindemann J. Numerical simulation ofhumidification and heating during inspiration within an adultnose. Rhinology. 2012. doi:10.4193/Rhino.11.231.

102. Kleven M. Numerical simulations and experimental studies fornasal delivery of vaccines and drugs. Thesis. NTNU. 2012.ISBN:978-82-471-3398-9.

103. Guo Y, Laube B, Dalby R. The effect of formulation varia-bles and breathing patterns on the site of nasal deposition inan anatomically correct model. Pharm Res. 2005;22(11):1871–8.

104. Kundoor V, Dalby RN. Assessment of nasal spray deposition in asilicone human nose model using a color based method. PharmRes. 2010;27(1):30–6.

105. Djupesland PG. A critical view on the use of nasal casts to assessnasal spray deposition patterns. ARS 2012. Poster #P-2. http://www.american-rhinologic.org/previous_meeting.

106. Rudman KL, O’Brien EK, Leopold DA. Radiographic distribu-tion of drops and sprays within the sinonasal cavities. Am JRhinol Allergy. 2011;25:94–7.

107. Benninger MS, Hadley JA, Osguthorpe JD, Marple BF, LeopoldDA, Derebery MJ, et al. Techniques of intranasal steroid use.Otolaryngol Head Neck Surg. 2004;130:5–24.

108. Merkus P, Ebbens FA, Muller B, Fokkens WJ. The ‘best method’of topical nasal drug delivery: comparison of seven techniques.Rhinology. 2006;44:102–7.

109. Al-Ghananeem AM, Sandefer EP, Doll WJ, Page RC, Chang Y,Digenis GA. Gamma scintigraphy for testing bioequivalence: acase study on two cromolyn sodium nasal spray preparations. IntJ Pharm. 2008;357:70–6.

110. McInnes FJ, O’Mahony B, Lindsay B, et al. Nasal residence ofinsulin containing lyophilised nasal insert formulations, usinggamma scintigraphy. Eur J Pharm Sci. 2007;31:25–31.

111. Newman SP, Moren F, Clarke SW. Deposition pattern of nasalsprays in man. Rhinology. 1987;26:111–20.

112. Darquenne C. Aerosol deposition in health and disease. J AerosolMed Pulm. 2012;25(3):140–7.

113. McLean JA, Bacon JR, Mathews KP, Thrall JH, Banas JM,Hedden J, Bayne NK. Deposition and clearance of mucora-dioactive aerosol on the nasal mucosa. Rhinology. 1984;22:65–75.

114. Harris AS, Ohlin M, Lethagen S, Nilsson IM. Effect of concen-tration and volume on nasal bioavailability and biological re-sponse to desmopressin. J Pharm Sci. 1988;77:337–9.

115. Costantino HR, Illum L, Brandt G, Johnson PH, Quaya SC.Intranasal delivery: physicochemical and therapeutic aspects. IntJ Pharma. 2007;337:1–24.

116. Bond W, Hardy JG, Wilson CG. Deposition and clearance ofnasal sprays. In: Aiache JM, Hirtz J, editors. Proceedings volumeI, Biopharmaceutics; 1985. pp. 93–8.

117. Hardy JG, Lee SW, Wilson CG. Intranasal drug delivery by sprayand drops. J Pharm Pharmacol. 1985;37:294–7.

118. Righton L. Bringing the patient’s voice to nasal drug delivery.ONdrugDelivery. www.ondrugdelivery.com. 2011;4–7.

62 Drug Deliv. and Transl. Res. (2013) 3:42–62